diff --git a/CHANGELOG.md b/CHANGELOG.md
index 6fb9152c6..a238977a3 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -12,6 +12,8 @@
 - Match scorers now return answers consistently even when there is no match.
 - Relocate tool related types into a new top-level `inspect_ai.tool` module (previous imports still work fow now, but result in a runtime deprecation warning)
 - Decouple tools entirely from solvers and task state (previously they had ways to interact with metadata, removing this coupling will enable tool use in lower level interactions with models). Accordingly, the `call_tools()` function now operates directly on messages rather than task state.
+- Use multiple scorers to score a task (by passing a list of scorers rather than a single scorer).
+- Support for complex scorers that return multiple scores and metrics (use a dictionary in the `@scorer` declaration to map metrics to the score value to apply the metrics to).
 
 ## v0.3.17 (25 June 2024)
 
diff --git a/src/inspect_ai/_cli/score.py b/src/inspect_ai/_cli/score.py
index 7f6367308..4575ffe07 100644
--- a/src/inspect_ai/_cli/score.py
+++ b/src/inspect_ai/_cli/score.py
@@ -89,6 +89,7 @@ async def score(
     # print results
     display().print(f"\n{eval_log.eval.task}")
     if eval_log.results:
-        for name, metric in eval_log.results.metrics.items():
-            display().print(f"{name}: {metric.value}")
+        for score in eval_log.results.scores:
+            for name, metric in score.metrics.items():
+                display().print(f"{name}: {metric.value}")
     display().print(f"log: {log_file}\n")
diff --git a/src/inspect_ai/_display/rich.py b/src/inspect_ai/_display/rich.py
index 62054c2d1..35e4d2f06 100644
--- a/src/inspect_ai/_display/rich.py
+++ b/src/inspect_ai/_display/rich.py
@@ -379,17 +379,19 @@ def task_interrupted(
 def task_results(results: EvalResults) -> tuple[RenderableType, RenderableType]:
     theme = rich_theme()
     output: dict[str, str] = {}
-    for name, metric in results.metrics.items():
-        value = (
-            "1.0"
-            if metric.value == 1
-            else (
-                str(metric.value)
-                if isinstance(metric.value, int)
-                else f"{metric.value:.3g}"
+    for score in results.scores:
+        for name, metric in score.metrics.items():
+            value = (
+                "1.0"
+                if metric.value == 1
+                else (
+                    str(metric.value)
+                    if isinstance(metric.value, int)
+                    else f"{metric.value:.3g}"
+                )
             )
-        )
-        output[name] = value
+            key = f"{score.name}/{name}" if len(results.scores) > 1 else name
+            output[key] = value
     metrics = f"[{theme.metric}]{task_dict(output, True)}[/{theme.metric}]"
 
     return (metrics, "")
diff --git a/src/inspect_ai/_eval/score.py b/src/inspect_ai/_eval/score.py
index db8330523..c2bf9ee24 100644
--- a/src/inspect_ai/_eval/score.py
+++ b/src/inspect_ai/_eval/score.py
@@ -8,6 +8,7 @@
 from inspect_ai._util.platform import platform_init
 from inspect_ai._util.registry import (
     registry_create,
+    registry_log_name,
 )
 from inspect_ai.log import (
     EvalLog,
@@ -22,12 +23,12 @@
 from .task.util import task_run_dir
 
 
-def score(log: EvalLog, scorer: Scorer) -> EvalLog:
+def score(log: EvalLog, scorers: Scorer | list[Scorer]) -> EvalLog:
     """Score an evaluation log.
 
     Args:
        log (EvalLog): Evaluation log.
-       scorer (Scorer): Scorer to apply to log
+       scorers (Scorer): List of Scorers to apply to log
        metrics: (list[Metric]): Additional metrics to compute
          (Scorer built-in metrics are always computed).
 
@@ -37,15 +38,18 @@ def score(log: EvalLog, scorer: Scorer) -> EvalLog:
     # standard platform init for top level entry points
     platform_init()
 
-    return asyncio.run(score_async(log, scorer))
+    # resolve scorers into a list
+    scorers = [scorers] if isinstance(scorers, Scorer) else scorers
 
+    return asyncio.run(score_async(log, scorers))
 
-async def score_async(log: EvalLog, scorer: Scorer) -> EvalLog:
+
+async def score_async(log: EvalLog, scorers: list[Scorer]) -> EvalLog:
     """Score an evaluation log.
 
     Args:
        log (EvalLog): Evaluation log.
-       scorer (Scorer): Scorer to apply to log
+       scorers (list[Scorer]): Scorers to apply to log
 
     Returns:
        Log with scores yielded by scorer.
@@ -78,23 +82,23 @@ def progress() -> None:
             p.update(1)
 
         tasks = [
-            run_score_task(state, Target(sample.target), scorer, progress)
+            run_score_task(state, Target(sample.target), scorers, progress)
             for (sample, state) in zip(log.samples, states)
         ]
 
         # do scoring
-        scores = await asyncio.gather(*tasks)
+        scores: list[dict[str, Score]] = await asyncio.gather(*tasks)
 
         # write them back (gather ensures that they come back in the same order)
         for index, score in enumerate(scores):
-            log.samples[index].score = score
+            log.samples[index].scores = score
 
         # collect metrics from EvalLog (they may overlap w/ the scorer metrics,
         # that will be taken care of in eval_results)
         log_metrics = metrics_from_log(log)
 
         # compute metrics
-        log.results = eval_results(scores, scorer, log_metrics)
+        log.results = eval_results(scores, scorers, log_metrics)
 
     return log
 
@@ -119,7 +123,7 @@ async def task_score(task: Task, log: EvalLog) -> EvalLog:
     display().print(f"Aggregating scores for task: {task_name}")
     if task.scorer and log.samples:
         log.results = eval_results(
-            [sample.score for sample in log.samples if isinstance(sample.score, Score)],
+            [sample.scores for sample in log.samples if sample.scores is not None],
             task.scorer,
             task.metrics,
         )
@@ -129,17 +133,25 @@ async def task_score(task: Task, log: EvalLog) -> EvalLog:
 async def run_score_task(
     state: TaskState,
     target: Target,
-    scorer: Scorer,
+    scorers: list[Scorer],
     progress: Callable[..., None],
-) -> Score:
-    result = await scorer(state, target)
+) -> dict[str, Score]:
+    results: dict[str, Score] = {}
+    for scorer in scorers:
+        result = await scorer(state, target)
+        results[registry_log_name(scorer)] = result
+
     progress()
-    return result
+    return results
 
 
 def metrics_from_log(log: EvalLog) -> list[Metric]:
     return (
-        [metric_from_log(metric) for metric in log.results.metrics.values()]
+        [
+            metric_from_log(metric)
+            for score in log.results.scores
+            for metric in score.metrics.values()
+        ]
         if log.results
         else []
     )
diff --git a/src/inspect_ai/_eval/task/log.py b/src/inspect_ai/_eval/task/log.py
index 110eb753c..a9c844c85 100644
--- a/src/inspect_ai/_eval/task/log.py
+++ b/src/inspect_ai/_eval/task/log.py
@@ -122,7 +122,7 @@ def log_sample(
         epoch: int,
         sample: Sample,
         state: TaskState,
-        score: Score | None,
+        scores: dict[str, Score] | None,
         flush: bool = False,
     ) -> None:
         # log
@@ -137,7 +137,7 @@ def log_sample(
                 metadata=state.metadata if state.metadata else {},
                 messages=state.messages,
                 output=state.output,
-                score=score,
+                scores=scores,
             ),
             flush,
         )
diff --git a/src/inspect_ai/_eval/task/results.py b/src/inspect_ai/_eval/task/results.py
index 193ea669f..1bbb33aa0 100644
--- a/src/inspect_ai/_eval/task/results.py
+++ b/src/inspect_ai/_eval/task/results.py
@@ -1,5 +1,6 @@
 import re
 from copy import deepcopy
+from typing import Any, cast
 
 from inspect_ai._util.registry import (
     registry_info,
@@ -10,42 +11,156 @@
 from inspect_ai.log import (
     EvalMetric,
     EvalResults,
-    EvalScorer,
+    EvalScore,
 )
 from inspect_ai.scorer import Metric, Score, Scorer
 from inspect_ai.scorer._scorer import SCORER_METRICS, scorer_metrics
 
 
 def eval_results(
-    scores: list[Score], scorer: Scorer | None, metrics: list[Metric] = []
+    scores: list[dict[str, Score]],
+    scorers: list[Scorer] | None,
+    metrics: list[Metric] = [],
 ) -> EvalResults:
     # record scorer
     results = EvalResults()
-    if scorer:
-        # extract non-metrics metadata
-        metadata = deepcopy(registry_info(scorer).metadata)
-        del metadata[SCORER_METRICS]
+    if scorers:
+        result_scores = []
+        for scorer in scorers:
+            # extract non-metrics metadata
+            metadata = deepcopy(registry_info(scorer).metadata)
+            del metadata[SCORER_METRICS]
 
+            # this scorer
+            scorer_name = registry_log_name(scorer)
+
+            # scores for this scorer
+            resolved_scores = [
+                score[scorer_name] for score in scores if scorer_name in score
+            ]
+
+            # Compute metrics for this scorer
+            targets = target_metrics(scorer, metrics)
+            if isinstance(targets, list):
+                # If there is a simple list of metrics
+                # just compute the metrics for this scorer
+                result_scores.extend(
+                    scorer_for_metrics(
+                        scorer_name=scorer_name,
+                        scorer=scorer,
+                        metadata=metadata,
+                        scores=resolved_scores,
+                        metrics=targets,
+                    )
+                )
+            else:
+                # If there is a dictionary of metrics, apply
+                # the metrics to the values within the scores
+                # (corresponding by key) and emit an EvalScorer for
+                # each key (which effectively creates multiple scorers
+                # by expanding a dictionary score value into multiple
+                # results with metrics)
+                result_scores.extend(
+                    scorers_from_metric_dict(
+                        scorer_name=scorer_name,
+                        scorer=scorer,
+                        metadata=metadata,
+                        scores=resolved_scores,
+                        metrics=targets,
+                    )
+                )
         # build results
-        results.scorer = EvalScorer(
-            name=registry_log_name(scorer),
+        results.scores = result_scores
+
+    return results
+
+
+def scorer_for_metrics(
+    scorer_name: str,
+    scorer: Scorer,
+    metadata: dict[str, Any],
+    scores: list[Score],
+    metrics: list[Metric],
+) -> list[EvalScore]:
+    results: list[EvalScore] = []
+    # we want to use simple names for metrics in the metrics dict
+    # (i.e. without package prefixes). we do this by getting the
+    # unqualified name, then appending a suffix if there are duplicates
+    # this keeps the code straightforward and intuitive for users
+    # programming against the log (e.g. metrics["accuracy"]) vs.
+    # metrics["pkgname/accuracy"])
+    list_metrics: dict[str, EvalMetric] = {}
+    for metric in metrics:
+        key = metrics_unique_key(
+            registry_unqualified_name(metric), list(list_metrics.keys())
+        )
+
+        list_metrics[key] = EvalMetric(
+            name=registry_log_name(metric),
+            value=cast(float, metric(scores)),
+        )
+
+    # build results
+    results.append(
+        EvalScore(
+            scorer=scorer_name,
+            name=scorer_name,
             params=registry_params(scorer),
             metadata=metadata if len(metadata.keys()) > 0 else None,
+            metrics=list_metrics,
         )
+    )
+    return results
+
 
-        # we want to use simple names for metrics in the metrics dict
-        # (i.e. without package prefixes). we do this by getting the
-        # unqualified name, then appending a suffix if there are duplicates
-        # this keeps the code straightforward and intuitive for users
-        # programming against the log (e.g. metrics["accuracy"]) vs.
-        # metrics["pkgname/accuracy"])
-        for metric in target_metrics(scorer, metrics):
-            key = metrics_unique_key(
-                registry_unqualified_name(metric), list(results.metrics.keys())
+def scorers_from_metric_dict(
+    scorer_name: str,
+    scorer: Scorer,
+    metadata: dict[str, Any],
+    scores: list[Score],
+    metrics: dict[str, list[Metric]],
+) -> list[EvalScore]:
+    results: list[EvalScore] = []
+    for metric_key, metric_list in metrics.items():
+        # filter scores to a list of scalars with the value of the metric name
+        metric_scores: list[Score] = []
+        for score in scores:
+            if isinstance(score.value, dict):
+                if metric_key in score.value:
+                    # Convert the score into a simple scalar value to apply metrics
+                    metric_score = deepcopy(score)
+                    metric_score.value = cast(float, score.value[metric_key])
+                    metric_scores.append(metric_score)
+                else:
+                    raise TypeError(
+                        f"key '{metric_key}' isn't present in the score value dictionary"
+                    )
+            else:
+                raise TypeError(
+                    "dictionary of metrics specific for a non-dictionary score"
+                )
+
+        result_metrics: dict[str, EvalMetric] = {}
+        for target_metric in metric_list:
+            # compute the metric value
+            metric_name = registry_log_name(target_metric)
+            result_metrics[metric_name] = EvalMetric(
+                name=metric_name,
+                value=cast(float, target_metric(metric_scores)),
             )
-            results.metrics[key] = EvalMetric(
-                name=registry_log_name(metric), value=metric(scores)
+
+        # create a scorer result for this metric
+        # TODO: What is there is separate simple scorer which has a name collision with
+        # a score created by this scorer
+        results.append(
+            EvalScore(
+                scorer=scorer_name,
+                name=metric_key,
+                params=registry_params(scorer),
+                metadata=metadata if len(metadata.keys()) > 0 else None,
+                metrics=result_metrics,
             )
+        )
     return results
 
 
@@ -64,14 +179,23 @@ def metrics_unique_key(key: str, existing: list[str]) -> str:
 
 
 # build a list of metrics (scorer built-in metrics + de-duplicated additional metrics)
-def target_metrics(scorer: Scorer, metrics: list[Metric]) -> list[Metric]:
-    target_metrics = scorer_metrics(scorer)
-    target_metrics_names = [registry_log_name(metric) for metric in target_metrics]
-    target_metrics.extend(
-        [
-            metric
-            for metric in metrics
-            if registry_log_name(metric) not in target_metrics_names
-        ]
-    )
-    return target_metrics
+def target_metrics(
+    scorer: Scorer, metrics: list[Metric]
+) -> list[Metric] | dict[str, list[Metric]]:
+    output_metrics = scorer_metrics(scorer)
+
+    if isinstance(output_metrics, dict):
+        if isinstance(metrics, dict):
+            output_metrics.update(metrics)
+        return output_metrics
+    else:
+        output_metrics_names = [registry_log_name(metric) for metric in output_metrics]
+        if isinstance(metrics, list):
+            output_metrics.extend(
+                [
+                    metric
+                    for metric in metrics
+                    if registry_log_name(metric) not in output_metrics_names
+                ]
+            )
+        return output_metrics
diff --git a/src/inspect_ai/_eval/task/run.py b/src/inspect_ai/_eval/task/run.py
index b5479b65e..67b16811b 100644
--- a/src/inspect_ai/_eval/task/run.py
+++ b/src/inspect_ai/_eval/task/run.py
@@ -141,7 +141,12 @@ async def task_run(
         else task.plan
     )
     score = score and task.scorer is not None
-    scorer: Scorer | None = task.scorer if (score and task.scorer) else None
+    scorers: list[Scorer] | None = task.scorer if (score and task.scorer) else None
+    scorer_profiles = (
+        [registry_log_name(scorer) for scorer in scorers if is_registry_object(scorer)]
+        if scorers is not None
+        else ["(none)"]
+    )
 
     # compute steps (steps = samples * steps in plan + 1 for scorer)
     steps = len(samples) * (
@@ -153,7 +158,7 @@ async def task_run(
         name=task.name,
         model=model_name,
         dataset=task.dataset.name or "(samples)",
-        scorer=(registry_log_name(scorer) if is_registry_object(scorer) else "(none)"),
+        scorer=", ".join(scorer_profiles),
         samples=len(samples),
         steps=steps,
         eval_config=config,
@@ -206,7 +211,7 @@ async def generate(
                         toolenv_cleanup=toolenv_cleanup,
                         plan=plan,
                         max_messages=config.max_messages,
-                        scorer=scorer,
+                        scorers=scorers,
                         generate=generate,
                         progress=progress,
                         logger=logger if log_samples else None,
@@ -221,11 +226,14 @@ async def generate(
                 scores = await asyncio.gather(*sample_coroutines)
 
             # compute and record metrics if we have scores
-            completed_scores = [score for score in scores if isinstance(score, Score)]
+            completed_scores = [
+                score_dict for score_dict in scores if isinstance(score_dict, dict)
+            ]
+
             if len(completed_scores) > 0:
                 results = eval_results(
                     scores=completed_scores,
-                    scorer=scorer,
+                    scorers=scorers,
                     metrics=task.metrics,
                 )
                 logger.log_results(results)
@@ -289,14 +297,14 @@ async def task_run_sample(
     toolenv_cleanup: bool,
     plan: Plan,
     max_messages: int | None,
-    scorer: Scorer | None,
+    scorers: list[Scorer] | None,
     generate: Generate,
     progress: Callable[..., None],
     logger: TaskLogger | None,
     log_images: bool,
     sample_source: EvalSampleSource | None,
     semaphore: asyncio.Semaphore | None,
-) -> Score | None:
+) -> dict[str, Score] | None:
     # if there is an existing sample then tick off its progress, log it, and return it
     if sample_source and sample.id is not None:
         previous_sample = sample_source(sample.id, state.epoch)
@@ -309,7 +317,7 @@ async def task_run_sample(
                 logger.log_event("sample", previous_sample, False)
 
             # return score
-            return previous_sample.score
+            return previous_sample.scores
 
     # use semaphore if provided
     semaphore_cm: asyncio.Semaphore | contextlib.AbstractAsyncContextManager[None] = (
@@ -356,7 +364,15 @@ async def task_run_sample(
                     )
 
         # score it
-        result = await scorer(state, Target(sample.target)) if scorer else None
+        results: dict[str, Score] = {}
+        if scorers:
+            for scorer in scorers:
+                scorer_name = registry_log_name(scorer)
+                score_result = (
+                    await scorer(state, Target(sample.target)) if scorer else None
+                )
+                if score_result is not None:
+                    results[scorer_name] = score_result
         progress()
 
         # log it
@@ -366,10 +382,10 @@ async def task_run_sample(
                 state = (await states_with_base64_images([state]))[0]
 
             # log the sample
-            logger.log_sample(state.epoch, sample, state, result, True)
+            logger.log_sample(state.epoch, sample, state, results, True)
 
         # return
-        return result
+        return results
 
 
 async def resolve_dataset(
diff --git a/src/inspect_ai/_eval/task/task.py b/src/inspect_ai/_eval/task/task.py
index c1129946f..6252b1bbf 100644
--- a/src/inspect_ai/_eval/task/task.py
+++ b/src/inspect_ai/_eval/task/task.py
@@ -25,7 +25,7 @@ class Task:
         dataset (Dataset | Sequence[Sample]): Dataset to evaluate
         plan: (Plan | Solver | list[Solver]): Default plan. If not specified
           defaults to generate(), a normal call to the model.
-        scorer: (Scorer | None): Scorer used to evaluate model output.
+        scorer: (Scorer | list[Scorer] | None): Scorer used to evaluate model output.
         metrics (list[Metric]): Additional metrics to compute beyond
           the base metrics provided by the scorer.
         config (GenerateConfig): Model generation config.
@@ -45,7 +45,7 @@ def __init__(
         self,
         dataset: Dataset | Sequence[Sample],
         plan: Plan | Solver | list[Solver] = generate(),
-        scorer: Scorer | None = None,
+        scorer: Scorer | list[Scorer] | None = None,
         metrics: list[Metric] = [],
         config: GenerateConfig = GenerateConfig(),
         tool_environment: str | tuple[str, str] | None = None,
@@ -58,7 +58,13 @@ def __init__(
             dataset if isinstance(dataset, Dataset) else MemoryDataset(list(dataset))
         )
         self.plan = plan if isinstance(plan, Plan) else Plan(plan)
-        self.scorer = scorer
+        self.scorer = (
+            scorer
+            if isinstance(scorer, list)
+            else [scorer]
+            if scorer is not None
+            else None
+        )
         self.metrics = metrics
         self.config = config
         self.tool_environment = (
diff --git a/src/inspect_ai/_view/view.py b/src/inspect_ai/_view/view.py
index 25844d189..5be2137f0 100644
--- a/src/inspect_ai/_view/view.py
+++ b/src/inspect_ai/_view/view.py
@@ -23,7 +23,7 @@
 )
 from inspect_ai._util.dotenv import init_dotenv
 from inspect_ai._util.error import exception_message
-from inspect_ai._util.file import FileSystem, file, filesystem
+from inspect_ai._util.file import FileSystem, filesystem
 from inspect_ai._util.http import InspectHTTPRequestHandler
 from inspect_ai.log._file import (
     eval_log_json,
@@ -180,9 +180,8 @@ def handle_log(self) -> None:
                     pass
 
             if contents is None:  # normal read
-                with file(path, "rb") as f:
-                    # read file and determine its length
-                    contents = f.read()
+                log = read_eval_log(path, header_only=False)
+                contents = eval_log_json(log).encode()
 
             # respond with the log
             length = len(contents)
diff --git a/src/inspect_ai/_view/www/App.css b/src/inspect_ai/_view/www/App.css
index 9984392c3..5e243f91f 100644
--- a/src/inspect_ai/_view/www/App.css
+++ b/src/inspect_ai/_view/www/App.css
@@ -4,6 +4,18 @@
   --bs-popover-max-width: 50%;
 }
 
+#app {
+  height: 100vh;
+  overflow-y: hidden;
+}
+
+.app-main-grid {
+  display: grid;
+  height: 100vh;
+  overflow-y: hidden;
+  grid-template-rows: minmax(65px, max-content) max-content 1fr;
+}
+
 .modal {
   --bs-modal-margin: 0.5rem;
 }
@@ -12,6 +24,10 @@
   --bs-backdrop-opacity: 0.4;
 }
 
+body[class^="vscode-"] .app-main-grid {
+  grid-template-rows: minmax(55px, max-content) max-content 1fr;
+}
+
 body[class^="vscode-"] {
   --bs-body-bg: var(--vscode-editor-background);
   --bs-card-bg: var(--vscode-editor-background);
@@ -57,21 +73,20 @@ body[class^="vscode-"] .modal-content {
   background-clip: unset;
 }
 
+body[class^="vscode-"] .multi-score-label {
+  margin-bottom: 5px;
+}
+
 body[class^="vscode-"] {
   min-width: 400px;
 }
 
 body[class^="vscode-"] .navbar-brand {
-  font-size: 0.9em;
-}
-
-body[class^="vscode-"] .navbar-brand > div {
-  margin-top: 0.3em;
+  font-size: 1em;
 }
 
-body[class^="vscode-"] .navbar-metrics {
-  margin-top: -0.4em;
-  font-size: 0.7em;
+body[class^="vscode-"] .navbar-brand .navbar-secondary-text {
+  font-size: 0.8em;
 }
 
 body[class^="vscode-"] .navbar #sidebarToggle > i.bi {
@@ -186,7 +201,6 @@ body[class^="vscode-"] .sidebar .list-group {
 :root {
   --bs-navbar-padding-y: 0;
   --bs-navbar-brand-padding-y: 0;
-  --navbar-height: 50px;
   --sidebar-width: 500px;
 }
 
@@ -199,7 +213,6 @@ body {
 .navbar {
   padding-top: 0;
   padding-bottom: 0;
-  height: var(--navbar-height);
   background-color: var(--bs-light);
 }
 
@@ -209,7 +222,12 @@ body {
 
 .navbar-brand {
   font-weight: 400;
-  font-size: 1.2em;
+  font-size: 1.4em;
+}
+
+.navbar-text {
+  padding-top: 0px;
+  padding-bottom: 0px;
 }
 
 #sidebarToggle > i.bi {
@@ -230,7 +248,6 @@ body {
 .workspace {
   display: flex;
   flex-direction: column;
-  height: calc(100vh - var(--navbar-height));
 }
 
 .workspace.full-screen {
@@ -548,7 +565,6 @@ table.table.table-sm td {
 .tab-tools > * {
   flex: 0 1 auto;
   height: 1.5rem;
-  max-width: 150px;
   margin-left: 0.5rem;
 }
 
diff --git a/src/inspect_ai/_view/www/App.mjs b/src/inspect_ai/_view/www/App.mjs
index 966e54f1d..cf4031769 100644
--- a/src/inspect_ai/_view/www/App.mjs
+++ b/src/inspect_ai/_view/www/App.mjs
@@ -107,8 +107,9 @@ export function App() {
           capabilities,
         );
         if (logContents) {
+          const log = logContents.parsed;
           setCurrentLog({
-            contents: logContents.parsed,
+            contents: log,
             name: targetLog.name,
             raw: logContents.raw,
           });
@@ -265,39 +266,39 @@ export function App() {
   // if there are no log files, then don't show sidebar
   const fullScreen = filteredLogs.files.length === 1 && !filteredLogs.log_dir;
 
-  const appEnvelope = [
-    html` <${Navbar}
-      file=${currentLog.name}
-      logs=${filteredLogs}
-      task=${currentLog.contents?.eval?.task}
-      model=${currentLog.contents?.eval?.model}
-      metrics=${currentLog.contents?.results?.metrics}
-      samples=${currentLog.contents?.samples}
-      status=${currentLog.contents?.status}
-      offcanvas=${offcanvas}
-    />`,
-  ];
-  if (!fullScreen) {
-    appEnvelope.push(html`
-      <${Sidebar}
-        logs=${filteredLogs}
-        logHeaders=${logHeaders}
-        loading=${headersLoading}
-        offcanvas=${offcanvas}
-        selectedIndex=${selected}
-        onSelectedIndexChanged=${(index) => {
-          setSelected(index);
-
-          // hide the sidebar offcanvas
-          var myOffcanvas = document.getElementById("sidebarOffCanvas");
-          var bsOffcanvas = bootstrap.Offcanvas.getInstance(myOffcanvas);
-          if (bsOffcanvas) {
-            bsOffcanvas.hide();
-          }
-        }}
-      />
-    `);
-  }
+  const navbar = html` <${Navbar}
+    file=${currentLog.name}
+    logs=${filteredLogs}
+    task=${currentLog.contents?.eval?.task}
+    model=${currentLog.contents?.eval?.model}
+    results=${currentLog.contents?.results}
+    samples=${currentLog.contents?.samples}
+    status=${currentLog.contents?.status}
+    offcanvas=${offcanvas}
+  />`;
+
+  const sidebar =
+    !fullScreen && currentLog.contents
+      ? html`
+          <${Sidebar}
+            logs=${filteredLogs}
+            logHeaders=${logHeaders}
+            loading=${headersLoading}
+            offcanvas=${offcanvas}
+            selectedIndex=${selected}
+            onSelectedIndexChanged=${(index) => {
+              setSelected(index);
+
+              // hide the sidebar offcanvas
+              var myOffcanvas = document.getElementById("sidebarOffCanvas");
+              var bsOffcanvas = bootstrap.Offcanvas.getInstance(myOffcanvas);
+              if (bsOffcanvas) {
+                bsOffcanvas.hide();
+              }
+            }}
+          />
+        `
+      : "";
 
   const workspace = useMemo(() => {
     if (status.error) {
@@ -317,10 +318,14 @@ export function App() {
     }
   }, [logs, currentLog, selected, fullScreen, offcanvas, status]);
 
+  const fullScreenClz = fullScreen ? " full-screen" : "";
+  const offcanvasClz = offcanvas ? " off-canvas" : "";
+
   return html`
     <${AppErrorBoundary}>
-    <div>
-      ${appEnvelope}
+    ${sidebar}
+    <div class="app-main-grid${fullScreenClz}${offcanvasClz}">
+      ${navbar}
       <${ProgressBar} animating=${status.loading} />
       ${workspace}
     </div>
diff --git a/src/inspect_ai/_view/www/log-schema.json b/src/inspect_ai/_view/www/log-schema.json
index 94c1cf5b0..fa8bf07ef 100644
--- a/src/inspect_ai/_view/www/log-schema.json
+++ b/src/inspect_ai/_view/www/log-schema.json
@@ -652,24 +652,13 @@
     },
     "EvalResults": {
       "properties": {
-        "scorer": {
-          "anyOf": [
-            {
-              "$ref": "#/$defs/EvalScorer"
-            },
-            {
-              "type": "null"
-            }
-          ],
-          "default": null
-        },
-        "metrics": {
-          "additionalProperties": {
-            "$ref": "#/$defs/EvalMetric"
+        "scores": {
+          "default": [],
+          "items": {
+            "$ref": "#/$defs/EvalScore"
           },
-          "default": {},
-          "title": "Metrics",
-          "type": "object"
+          "title": "Scores",
+          "type": "array"
         },
         "metadata": {
           "anyOf": [
@@ -686,7 +675,7 @@
       },
       "title": "EvalResults",
       "type": "object",
-      "required": ["scorer", "metrics", "metadata"],
+      "required": ["scores", "metadata"],
       "additionalProperties": false
     },
     "EvalRevision": {
@@ -807,16 +796,20 @@
         "output": {
           "$ref": "#/$defs/ModelOutput"
         },
-        "score": {
+        "scores": {
           "anyOf": [
             {
-              "$ref": "#/$defs/Score"
+              "additionalProperties": {
+                "$ref": "#/$defs/Score"
+              },
+              "type": "object"
             },
             {
               "type": "null"
             }
           ],
-          "default": null
+          "default": null,
+          "title": "Scores"
         },
         "metadata": {
           "title": "Metadata",
@@ -831,24 +824,36 @@
         "target",
         "messages",
         "output",
-        "score",
+        "scores",
         "metadata"
       ],
       "title": "EvalSample",
       "type": "object",
       "additionalProperties": false
     },
-    "EvalScorer": {
+    "EvalScore": {
       "properties": {
         "name": {
           "title": "Name",
           "type": "string"
         },
+        "scorer": {
+          "title": "Scorer",
+          "type": "string"
+        },
         "params": {
           "default": {},
           "title": "Params",
           "type": "object"
         },
+        "metrics": {
+          "additionalProperties": {
+            "$ref": "#/$defs/EvalMetric"
+          },
+          "default": [],
+          "title": "Metrics",
+          "type": "object"
+        },
         "metadata": {
           "anyOf": [
             {
@@ -862,8 +867,8 @@
           "title": "Metadata"
         }
       },
-      "required": ["name", "params", "metadata"],
-      "title": "EvalScorer",
+      "required": ["name", "scorer", "params", "metrics", "metadata"],
+      "title": "EvalScore",
       "type": "object",
       "additionalProperties": false
     },
@@ -1633,7 +1638,7 @@
   },
   "properties": {
     "version": {
-      "default": 1,
+      "default": 2,
       "title": "Version",
       "type": "integer"
     },
diff --git a/src/inspect_ai/_view/www/log.d.ts b/src/inspect_ai/_view/www/log.d.ts
index 559958154..cde6dc97b 100644
--- a/src/inspect_ai/_view/www/log.d.ts
+++ b/src/inspect_ai/_view/www/log.d.ts
@@ -57,10 +57,12 @@ export type NumChoices = number | null;
 export type Logprobs = boolean | null;
 export type TopLogprobs = number | null;
 export type Name2 = string;
-export type Metadata1 = {} | null;
+export type Scorer = string;
 export type Name3 = string;
 export type Value = number;
+export type Metadata1 = {} | null;
 export type Metadata2 = {} | null;
+export type Scores = EvalScore[];
 export type Metadata3 = {} | null;
 export type StartedAt = string;
 export type CompletedAt = string;
@@ -131,14 +133,16 @@ export type Bytes1 = number[] | null;
 export type Content4 = Logprob[];
 export type Choices1 = ChatCompletionChoice[];
 export type Error = string | null;
+export type Scores1 = {
+  [k: string]: Score;
+} | null;
 export type Value1 =
   | string
   | number
-  | number
   | boolean
-  | (string | number | number | boolean)[]
+  | (string | number | boolean)[]
   | {
-      [k: string]: string | number | number | boolean | null;
+      [k: string]: string | number | boolean | null;
     };
 export type Answer = string | null;
 export type Explanation = string | null;
@@ -248,14 +252,15 @@ export interface GenerateConfig {
   top_logprobs: TopLogprobs;
 }
 export interface EvalResults {
-  scorer: EvalScorer | null;
-  metrics: Metrics;
+  scores: Scores;
   metadata: Metadata3;
 }
-export interface EvalScorer {
+export interface EvalScore {
   name: Name2;
+  scorer: Scorer;
   params: Params1;
-  metadata: Metadata1;
+  metrics: Metrics;
+  metadata: Metadata2;
 }
 export interface Params1 {}
 export interface Metrics {
@@ -265,7 +270,7 @@ export interface EvalMetric {
   name: Name3;
   value: Value;
   options: Options;
-  metadata: Metadata2;
+  metadata: Metadata1;
 }
 export interface Options {}
 export interface EvalStats {
@@ -294,7 +299,7 @@ export interface EvalSample {
   target: Target;
   messages: Messages;
   output: ModelOutput;
-  score: Score | null;
+  scores: Scores1;
   metadata: Metadata5;
 }
 export interface ChatMessageSystem {
diff --git a/src/inspect_ai/_view/www/src/Constants.mjs b/src/inspect_ai/_view/www/src/Constants.mjs
index bdbadaad5..5956e614f 100644
--- a/src/inspect_ai/_view/www/src/Constants.mjs
+++ b/src/inspect_ai/_view/www/src/Constants.mjs
@@ -29,6 +29,8 @@ export const icons = {
     critical: "bi bi-fire",
   },
   menu: "bi bi-list",
+  messages: "bi bi-chat-right-text",
+  metadata: "bi bi-table",
   model: "bi bi-cpu",
   "toggle-right": "bi bi-chevron-right",
   more: "bi bi-zoom-in",
diff --git a/src/inspect_ai/_view/www/src/components/CopyButton.mjs b/src/inspect_ai/_view/www/src/components/CopyButton.mjs
index 5641a5d4b..413da8928 100644
--- a/src/inspect_ai/_view/www/src/components/CopyButton.mjs
+++ b/src/inspect_ai/_view/www/src/components/CopyButton.mjs
@@ -4,7 +4,12 @@ import { icons } from "../Constants.mjs";
 export const CopyButton = ({ value }) => {
   return html`<button
     class="copy-button"
-    style=${{ border: "none", backgroundColor: "inherit", opacity: "0.5" }}
+    style=${{
+      border: "none",
+      backgroundColor: "inherit",
+      opacity: "0.5",
+      paddingTop: "0px",
+    }}
     data-clipboard-text=${value}
     onclick=${(e) => {
       const iEl = e.target;
diff --git a/src/inspect_ai/_view/www/src/components/ProgressBar.mjs b/src/inspect_ai/_view/www/src/components/ProgressBar.mjs
index 5c61d7fb2..dc72aa742 100644
--- a/src/inspect_ai/_view/www/src/components/ProgressBar.mjs
+++ b/src/inspect_ai/_view/www/src/components/ProgressBar.mjs
@@ -18,8 +18,7 @@ export const ProgressBar = ({ style, animating }) => {
 
   const progressContainerStyle = {
     width: "100%",
-    height: "6px",
-    marginBottom: "-6px",
+    height: "4px",
     background: "none",
   };
 
diff --git a/src/inspect_ai/_view/www/src/components/TabSet.mjs b/src/inspect_ai/_view/www/src/components/TabSet.mjs
index 695242fbd..f13f2275b 100644
--- a/src/inspect_ai/_view/www/src/components/TabSet.mjs
+++ b/src/inspect_ai/_view/www/src/components/TabSet.mjs
@@ -100,6 +100,12 @@ const Tab = ({ type, tab, index, style }) => {
           },
         }}
       >
+        ${tab.props.icon
+          ? html`<i
+              class="${tab.props.icon}"
+              style=${{ marginRight: "0.5em" }}
+            ></i>`
+          : ""}
         ${tab.props.title}
       </button>
     </li>
diff --git a/src/inspect_ai/_view/www/src/navbar/Navbar.mjs b/src/inspect_ai/_view/www/src/navbar/Navbar.mjs
index c73ded971..d41cfaaae 100644
--- a/src/inspect_ai/_view/www/src/navbar/Navbar.mjs
+++ b/src/inspect_ai/_view/www/src/navbar/Navbar.mjs
@@ -13,7 +13,7 @@ export const Navbar = ({
   model,
   status,
   samples,
-  metrics,
+  results,
   offcanvas,
 }) => {
   const toggleOffCanClass = offcanvas ? "" : " d-md-none";
@@ -21,7 +21,7 @@ export const Navbar = ({
 
   let statusPanel;
   if (status === "success") {
-    statusPanel = html`<${ResultsPanel} results="${metrics}" />`;
+    statusPanel = html`<${ResultsPanel} results="${results}" />`;
   } else if (status === "cancelled") {
     statusPanel = html`<${CanceledPanel}
       sampleCount=${samples?.length || 0}
@@ -34,7 +34,12 @@ export const Navbar = ({
   const navbarContents = logFileName
     ? html` <div
           class="navbar-brand navbar-text mb-0"
-          style=${{ display: "flex", paddingTop: 0, marginLeft: "0.5rem" }}
+          style=${{
+            display: "flex",
+            paddingTop: 0,
+            marginLeft: "0.5rem",
+            minWidth: "350px",
+          }}
         >
           ${logs.files.length > 1 || logs.log_dir
             ? html`<button
@@ -46,21 +51,25 @@ export const Navbar = ({
                 aria-controls="sidebarOffCanvas"
                 style=${{
                   padding: "0rem 0.1rem 0.1rem 0rem",
-                  marginTop: "-8px",
-                  marginRight: "0.2rem",
-                  lineHeight: "16px",
+                  display: "flex",
                 }}
               >
                 <i class=${icons.menu}></i>
               </button> `
             : ""}
-          <div style=${{ display: "flex", flexDirection: "column" }}>
+          <div
+            style=${{
+              display: "flex",
+              flexDirection: "column",
+              marginLeft: "0.2rem",
+            }}
+          >
             <div
               style=${{
-                marginTop: "0.5rem",
+                marginTop: "0.1rem",
                 display: "grid",
                 gridTemplateColumns:
-                  "minmax(50px,max-content) minmax(100px, max-content)",
+                  "minmax(30px,max-content) minmax(100px, max-content)",
               }}
             >
               <div
@@ -88,13 +97,16 @@ export const Navbar = ({
             <div
               style=${{
                 opacity: "0.7",
+                marginTop: "0.1rem",
                 paddingBottom: 0,
-                fontSize: "0.7rem",
+                fontSize: "0.9rem",
+                fontWeight: "300",
                 display: "grid",
                 gridTemplateColumns: "minmax(0,max-content) max-content",
               }}
             >
               <div
+                class="navbar-secondary-text"
                 style=${{
                   ...sharedStyles.wrapText(),
                 }}
@@ -111,6 +123,7 @@ export const Navbar = ({
           style=${{
             justifyContent: "end",
             marginRight: "1em",
+            marginBottom: "0",
           }}
         >
           ${statusPanel}
@@ -128,7 +141,7 @@ export const Navbar = ({
       <div
         style=${{
           display: "grid",
-          gridTemplateColumns: "1fr max-content",
+          gridTemplateColumns: "1fr auto",
           width: "100%",
         }}
       >
@@ -171,43 +184,109 @@ const RunningPanel = () => {
 
 const ResultsPanel = ({ results }) => {
   // Map the scores into a list of key/values
-  const metrics = results
-    ? Object.keys(results).map((key) => {
-        return { name: key, value: results[key].value };
-      })
-    : [];
+  const scorers = {};
+  results.scores.map((score) => {
+    scorers[score.name] = Object.keys(score.metrics).map((key) => {
+      return { name: key, value: score.metrics[key].value };
+    });
+  });
 
-  return html`<div
-    style=${{
-      display: "flex",
-      flexDirection: "row",
-      flexWrap: "wrap",
-      justifyContent: "end",
-    }}
-  >
-    ${metrics.map((metric, i) => {
-      return html`<div style=${{ paddingLeft: i === 0 ? "0" : "1em" }}>
-        <div
-          style=${{
-            fontSize: "0.7rem",
-            fontWeight: "200",
-            textAlign: "center",
-            marginBottom: "-0.3rem",
-            paddingTop: "0.3rem",
-          }}
-        >
-          ${metric.name}
-        </div>
-        <div
-          style=${{
-            fontSize: "1.5rem",
-            fontWeight: "500",
-            textAlign: "center",
-          }}
-        >
-          ${formatPrettyDecimal(metric.value)}
-        </div>
-      </div>`;
-    })}
+  if (results.scores.length === 1) {
+    const metrics = Object.values(scorers)[0];
+    return html`<div
+      style=${{
+        display: "flex",
+        flexDirection: "row",
+        flexWrap: "wrap",
+        justifyContent: "end",
+        height: "100%",
+        alignItems: "center",
+      }}
+    >
+      ${metrics.map((metric, i) => {
+        return html`<${VerticalMetric} metric=${metric} isFirst=${i === 0} />`;
+      })}
+    </div>`;
+  } else {
+    return html`<div
+      style=${{
+        display: "flex",
+        flexDirection: "row",
+        flexWrap: "wrap",
+        justifyContent: "end",
+        height: "100%",
+        alignItems: "center",
+        marginTop: "0.2rem",
+        paddingBottom: "0.4rem",
+        rowGap: "1em",
+      }}
+    >
+      ${results.scores.map((score, index) => {
+        return html`<${MultiScorerMetric}
+          scorer=${score}
+          isFirst=${index === 0}
+        />`;
+      })}
+    </div>`;
+  }
+};
+
+const VerticalMetric = ({ metric, isFirst }) => {
+  return html`<div style=${{ paddingLeft: isFirst ? "0" : "1em" }}>
+    <div
+      style=${{
+        fontSize: "0.8rem",
+        fontWeight: "200",
+        textAlign: "center",
+        marginBottom: "-0.3rem",
+        paddingTop: "0.3rem",
+      }}
+    >
+      ${metric.name}
+    </div>
+    <div
+      style=${{
+        fontSize: "1.5rem",
+        fontWeight: "500",
+        textAlign: "center",
+      }}
+    >
+      ${formatPrettyDecimal(metric.value)}
+    </div>
+  </div>`;
+};
+
+const MultiScorerMetric = ({ scorer, isFirst }) => {
+  const baseFontSize = Object.keys(scorer.metrics).length === 1 ? 0.9 : 0.7;
+  return html`<div style=${{ paddingLeft: isFirst ? "0" : "1.5em" }}>
+    <div
+      style=${{
+        fontSize: `${baseFontSize}rem`,
+        fontWeight: "200",
+        textAlign: "center",
+        borderBottom: "solid var(--bs-border-color) 1px",
+        textTransform: "uppercase",
+      }}
+      class="multi-score-label"
+    >
+      ${scorer.name}
+    </div>
+    <div
+      style=${{
+        display: "grid",
+        gridTemplateColumns: "auto auto",
+        gridColumnGap: "0.3rem",
+        gridRowGap: "0",
+        fontSize: `${baseFontSize + 0.1}rem`,
+      }}
+    >
+      ${Object.keys(scorer.metrics).map((key) => {
+        const metric = scorer.metrics[key];
+        return html` <div>${metric.name}</div>
+          <div style=${{ fontWeight: "600" }}>
+            ${formatPrettyDecimal(metric.value)}
+          </div>`;
+      })}
+    </div>
   </div>`;
 };
diff --git a/src/inspect_ai/_view/www/src/samples/SampleDisplay.mjs b/src/inspect_ai/_view/www/src/samples/SampleDisplay.mjs
index b4b9151cf..85a4ca3b4 100644
--- a/src/inspect_ai/_view/www/src/samples/SampleDisplay.mjs
+++ b/src/inspect_ai/_view/www/src/samples/SampleDisplay.mjs
@@ -7,12 +7,8 @@ import { TabSet, TabPanel } from "../components/TabSet.mjs";
 
 import { inputString } from "../utils/Format.mjs";
 
-import { sharedStyles } from "../Constants.mjs";
-import {
-  arrayToString,
-  shortenCompletion,
-  answerForSample,
-} from "../utils/Format.mjs";
+import { icons, sharedStyles } from "../Constants.mjs";
+import { arrayToString, shortenCompletion } from "../utils/Format.mjs";
 
 import { SampleScoreView } from "./SampleScoreView.mjs";
 import { MarkdownDiv } from "../components/MarkdownDiv.mjs";
@@ -67,25 +63,45 @@ export const SampleDisplay = ({
   // The core tabs
   const tabs = [
     html`
-    <${TabPanel} id=${msgTabId} title="Messages" onSelected=${onSelectedTab} selected=${
+    <${TabPanel} id=${msgTabId} title="Messages" icon=${icons.messages} onSelected=${onSelectedTab} selected=${
       selectedTab === msgTabId || selectedTab === undefined
     }>
       <${ChatView} key=${`${baseId}-chat`} id=${`${baseId}-chat`} messages=${
         sample.messages
       }/>
     </${TabPanel}>`,
-    html`
-    <${TabPanel} id=${scoringTabId} title="Scoring" onSelected=${onSelectedTab} selected=${
-      selectedTab === scoringTabId
-    }>
-      <${SampleScoreView}
-        sample=${sample}
-        context=${context}
-        sampleDescriptor=${sampleDescriptor}
-      />
-    </${TabPanel}>`,
   ];
 
+  const scorerNames = Object.keys(sample.scores);
+  if (scorerNames.length === 1) {
+    tabs.push(html`
+      <${TabPanel} id=${scoringTabId} title="Scoring" icon=${icons.scorer} onSelected=${onSelectedTab} selected=${
+        selectedTab === scoringTabId
+      }>
+        <${SampleScoreView}
+          sample=${sample}
+          context=${context}
+          sampleDescriptor=${sampleDescriptor}
+          scorer=${Object.keys(sample.scores)[0]}
+        />
+      </${TabPanel}>`);
+  } else {
+    for (const scorer of Object.keys(sample.scores)) {
+      const tabId = `score-${scorer}`;
+      tabs.push(html`
+        <${TabPanel} id="${tabId}" title="${scorer}" icon=${icons.scorer} onSelected=${onSelectedTab} selected=${
+          selectedTab === tabId
+        }>
+          <${SampleScoreView}
+            sample=${sample}
+            context=${context}
+            sampleDescriptor=${sampleDescriptor}
+            scorer=${scorer}
+          />
+        </${TabPanel}>`);
+    }
+  }
+
   const sampleMetadatas = metadataViewsForSample(baseId, sample, context);
   if (sampleMetadatas.length > 0) {
     tabs.push(
@@ -93,6 +109,7 @@ export const SampleDisplay = ({
       <${TabPanel} 
           id=${metdataTabId} 
           title="Metadata" 
+          icon=${icons.metadata}
           onSelected=${onSelectedTab} 
           selected=${selectedTab === metdataTabId}>
         ${sampleMetadatas}
@@ -201,7 +218,9 @@ const SampleSummary = ({ id, sample, sampleDescriptor }) => {
     });
   }
 
-  const fullAnswer = sample ? answerForSample(sample) : undefined;
+  const fullAnswer = sample
+    ? sampleDescriptor.selectedScorer(sample).answer()
+    : undefined;
   if (fullAnswer) {
     columns.push({
       label: "Answer",
@@ -219,11 +238,7 @@ const SampleSummary = ({ id, sample, sampleDescriptor }) => {
 
   columns.push({
     label: "Score",
-    value: sampleDescriptor?.scoreDescriptor.render
-      ? sampleDescriptor.scoreDescriptor.render(sample?.score?.value)
-      : sample?.score?.value === null
-        ? "null"
-        : sample?.score?.value,
+    value: sampleDescriptor?.selectedScore(sample).render(),
     size: "minmax(2em, auto)",
     center: true,
   });
diff --git a/src/inspect_ai/_view/www/src/samples/SampleList.mjs b/src/inspect_ai/_view/www/src/samples/SampleList.mjs
index fc05e1d5d..3cbb050a9 100644
--- a/src/inspect_ai/_view/www/src/samples/SampleList.mjs
+++ b/src/inspect_ai/_view/www/src/samples/SampleList.mjs
@@ -4,11 +4,7 @@ import { useEffect, useMemo } from "preact/hooks";
 import { sharedStyles } from "../Constants.mjs";
 import { MarkdownDiv } from "../components/MarkdownDiv.mjs";
 
-import {
-  shortenCompletion,
-  arrayToString,
-  answerForSample,
-} from "../utils/Format.mjs";
+import { shortenCompletion, arrayToString } from "../utils/Format.mjs";
 import { EmptyPanel } from "../components/EmptyPanel.mjs";
 import { VirtualList } from "../components/VirtualList.mjs";
 import { inputString } from "../utils/Format.mjs";
@@ -25,6 +21,7 @@ export const SampleList = (props) => {
     style,
     selectedIndex,
     setSelectedIndex,
+    selectedScore,
     nextSample,
     prevSample,
     showSample,
@@ -97,6 +94,7 @@ export const SampleList = (props) => {
           sampleDescriptor=${sampleDescriptor}
           selected=${selectedIndex === index}
           setSelected=${setSelectedIndex}
+          selectedScore=${selectedScore}
           showSample=${showSample}
         />
       `;
@@ -268,7 +266,9 @@ const SampleRow = ({
         ${sample
           ? html`
               <${MarkdownDiv}
-                markdown=${shortenCompletion(answerForSample(sample))}
+                markdown=${shortenCompletion(
+                  sampleDescriptor.selectedScorer(sample).answer(),
+                )}
                 style=${{ paddingLeft: "0" }}
                 class="no-last-para-padding"
               />
@@ -283,11 +283,7 @@ const SampleRow = ({
           display: "flex",
         }}
       >
-        ${sampleDescriptor?.scoreDescriptor.render
-          ? sampleDescriptor.scoreDescriptor.render(sample?.score?.value)
-          : sample?.score?.value === null
-            ? "null"
-            : sample?.score?.value}
+        ${sampleDescriptor?.selectedScore(sample).render()}
       </div>
     </div>
   `;
diff --git a/src/inspect_ai/_view/www/src/samples/SampleScoreView.mjs b/src/inspect_ai/_view/www/src/samples/SampleScoreView.mjs
index 11cb14011..821944840 100644
--- a/src/inspect_ai/_view/www/src/samples/SampleScoreView.mjs
+++ b/src/inspect_ai/_view/www/src/samples/SampleScoreView.mjs
@@ -2,10 +2,10 @@ import { html } from "htm/preact";
 import {
   arrayToString,
   shortenCompletion,
-  answerForSample,
   inputString,
 } from "../utils/Format.mjs";
 import { MarkdownDiv } from "../components/MarkdownDiv.mjs";
+import { SampleScores } from "./SampleScores.mjs";
 
 const labelStyle = {
   paddingRight: "2em",
@@ -13,7 +13,12 @@ const labelStyle = {
   paddingBottom: "0",
 };
 
-export const SampleScoreView = ({ sample, sampleDescriptor, style }) => {
+export const SampleScoreView = ({
+  sample,
+  sampleDescriptor,
+  style,
+  scorer,
+}) => {
   const scoreInput = [inputString(sample.input)];
   if (sample.choices && sample.choices.length > 0) {
     scoreInput.push("");
@@ -24,6 +29,10 @@ export const SampleScoreView = ({ sample, sampleDescriptor, style }) => {
     );
   }
 
+  const scorerDescriptor = sampleDescriptor.scorer(sample, scorer);
+  const explanation = scorerDescriptor.explanation() || "(No Explanation)";
+  const answer = scorerDescriptor.answer();
+
   return html`
     <div
       class="container-fluid"
@@ -54,7 +63,13 @@ export const SampleScoreView = ({ sample, sampleDescriptor, style }) => {
         </thead>
         <tbody style=${{ borderBottomColor: "#00000000" }}>
           <tr>
-            <td style=${{ paddingRight: "2em", paddingLeft: "0" }}>
+            <td
+              style=${{
+                paddingRight: "2em",
+                paddingLeft: "0",
+                paddingTop: "0",
+              }}
+            >
               <${MarkdownDiv}
                 markdown=${arrayToString(
                   arrayToString(sample?.target || "none"),
@@ -63,26 +78,25 @@ export const SampleScoreView = ({ sample, sampleDescriptor, style }) => {
                 class="no-last-para-padding"
               />
             </td>
-            <td>
+            <td style=${{ paddingTop: "0" }}>
               <${MarkdownDiv}
                 class="no-last-para-padding"
-                markdown=${shortenCompletion(answerForSample(sample))}
+                markdown=${shortenCompletion(answer)}
                 style=${{ paddingLeft: "0" }}
               />
             </td>
-            <td style=${{ paddingLeft: "2em" }}>
-              ${sampleDescriptor?.scoreDescriptor.render
-                ? sampleDescriptor.scoreDescriptor.render(sample?.score?.value)
-                : sample?.score?.value === null
-                  ? "null"
-                  : sample?.score?.value}
+            <td style=${{ paddingLeft: "2em", paddingTop: "0" }}>
+              <${SampleScores}
+                sample=${sample}
+                sampleDescriptor=${sampleDescriptor}
+                scorer=${scorer}
+              />
             </td>
           </tr>
         </tbody>
       </table>
 
-      ${sample?.score?.explanation &&
-      sample?.score?.explanation !== answerForSample(sample)
+      ${explanation && explanation !== answer
         ? html`
         <table class="table" style=${{ width: "100%", marginBottom: "0" }}>
               <thead>
@@ -95,9 +109,7 @@ export const SampleScoreView = ({ sample, sampleDescriptor, style }) => {
               </thead>
               <tbody>
                 <td style=${{ paddingLeft: "0" }}>
-                  <${MarkdownDiv} markdown=${arrayToString(
-                    sample?.score?.explanation,
-                  )} style=${{ paddingLeft: "0" }} class="no-last-para-padding"/>
+                  <${MarkdownDiv} markdown=${arrayToString(explanation)} style=${{ paddingLeft: "0" }} class="no-last-para-padding"/>
                 </td>
               </tbody>
             </table
diff --git a/src/inspect_ai/_view/www/src/samples/SampleScores.mjs b/src/inspect_ai/_view/www/src/samples/SampleScores.mjs
new file mode 100644
index 000000000..11e8d8725
--- /dev/null
+++ b/src/inspect_ai/_view/www/src/samples/SampleScores.mjs
@@ -0,0 +1,25 @@
+import { html } from "htm/preact";
+
+export const SampleScores = ({ sample, sampleDescriptor, scorer }) => {
+  const scores = scorer
+    ? sampleDescriptor.scorer(sample, scorer).scores()
+    : sampleDescriptor.selectedScorer(sample).scores();
+
+  if (scores.length === 1) {
+    return scores[0].rendered();
+  } else {
+    const rows = scores.map((score) => {
+      return html` <div style=${{ opacity: "0.7" }}>${score.name}</div>
+        <div>${score.rendered()}</div>`;
+    });
+    return html`<div
+      style=${{
+        display: "grid",
+        gridTemplateColumns: "max-content max-content",
+        columnGap: "1em",
+      }}
+    >
+      ${rows}
+    </div>`;
+  }
+};
diff --git a/src/inspect_ai/_view/www/src/samples/SamplesDescriptor.mjs b/src/inspect_ai/_view/www/src/samples/SamplesDescriptor.mjs
index 88262a4cd..b79cf02c5 100644
--- a/src/inspect_ai/_view/www/src/samples/SamplesDescriptor.mjs
+++ b/src/inspect_ai/_view/www/src/samples/SamplesDescriptor.mjs
@@ -5,7 +5,6 @@ import {
   formatDecimalNoTrailingZeroes,
   inputString,
   arrayToString,
-  answerForSample,
 } from "../utils/Format.mjs";
 import { RenderedContent } from "../components/RenderedContent.mjs";
 import { isNumeric } from "../utils/Type.mjs";
@@ -22,16 +21,99 @@ export const kScoreTypeOther = "other";
 export const kScoreTypeObject = "object";
 export const kScoreTypeBoolean = "boolean";
 
-export const samplesDescriptor = (samples, epochs, context) => {
+export const samplesDescriptor = (
+  selectedScore,
+  scorers,
+  samples,
+  epochs,
+  context,
+) => {
   if (!samples) {
     return undefined;
   }
 
+  const score = (sample, scorer = selectedScore?.scorer) => {
+    if (sample.scores[scorer]) {
+      return sample.scores[scorer];
+    } else {
+      return undefined;
+    }
+  };
+
+  // function for retrieving the sample score value
+  const scoreValue = (sample) => {
+    // no scores, no value
+    if (Object.keys(sample.scores).length === 0 || !selectedScore) {
+      return undefined;
+    }
+
+    if (
+      selectedScore.scorer !== selectedScore.name &&
+      sample.scores[selectedScore.scorer] &&
+      sample.scores[selectedScore.scorer].value
+    ) {
+      return sample.scores[selectedScore.scorer].value[selectedScore.name];
+    } else if (sample.scores[selectedScore.name]) {
+      return sample.scores[selectedScore.name].value;
+    } else {
+      return undefined;
+    }
+  };
+
+  // Retrieve the answer for a sample
+  const scoreAnswer = (sample, scorer) => {
+    if (sample) {
+      const sampleScore = score(sample, scorer);
+      if (sampleScore && sampleScore.answer) {
+        return sampleScore.answer;
+      } else if (sample.output.choices && sample.output.choices.length > 0) {
+        const content = sample.output.choices[0].message.content;
+        if (typeof content === "string") {
+          return content;
+        } else {
+          // TODO: Support image completions.
+          return content.length > 0 ? content[0].text : "";
+        }
+      }
+    } else {
+      return undefined;
+    }
+  };
+
+  const scoreExplanation = (sample, scorer) => {
+    if (sample) {
+      const sampleScore = score(sample, scorer);
+      if (sampleScore && sampleScore.explanation) {
+        return sampleScore.explanation;
+      }
+    }
+    return undefined;
+  };
+
   const uniqScoreValues = [
     ...new Set(
       samples
-        .filter((sample) => !!sample.score)
-        .map((sample) => sample.score.value)
+        .filter((sample) => !!sample.scores)
+        .filter((sample) => {
+          // There is no selected scorer, so include this value
+          if (!selectedScore) {
+            return true;
+          }
+
+          if (selectedScore.scorer !== selectedScore.name) {
+            return (
+              Object.keys(sample.scores).includes(selectedScore.scorer) &&
+              Object.keys(sample.scores[selectedScore.scorer].value).includes(
+                selectedScore.name,
+              )
+            );
+          } else {
+            return Object.keys(sample.scores).includes(selectedScore.name);
+          }
+        })
+        .map((sample) => {
+          return scoreValue(sample);
+        })
         .filter((value) => {
           return value !== null;
         }),
@@ -58,10 +140,7 @@ export const samplesDescriptor = (samples, epochs, context) => {
     (previous, current) => {
       previous[0] = Math.max(previous[0], inputString(current.input).length);
       previous[1] = Math.max(previous[1], arrayToString(current.target).length);
-      previous[2] = Math.max(
-        previous[2],
-        answerForSample(current)?.length || 0,
-      );
+      previous[2] = Math.max(previous[2], scoreAnswer(current)?.length || 0);
       return previous;
     },
     [0, 0, 0],
@@ -74,7 +153,103 @@ export const samplesDescriptor = (samples, epochs, context) => {
     target: sizes[1] / base,
     answer: sizes[2] / base,
   };
-  return { scoreDescriptor, epochs, messageShape };
+
+  const scoreRendered = (sample) => {
+    const score = scoreValue(sample);
+    if (score === null || score === "undefined") {
+      return "null";
+    } else if (scoreDescriptor.render) {
+      return scoreDescriptor.render(score);
+    } else {
+      return score;
+    }
+  };
+
+  const scorerDescriptor = (sample, scorer) => {
+    return {
+      explanation: () => {
+        return scoreExplanation(sample, scorer);
+      },
+      answer: () => {
+        return scoreAnswer(sample, scorer);
+      },
+      scores: () => {
+        if (!sample || !sample.scores) {
+          return [];
+        }
+
+        // Make a list of all the valid score names (this is
+        // used to distinguish between dictionaries that contain
+        // scores that should be treated as standlone scores and
+        // dictionaries that just contain random values, which is allowed)
+        const scoreNames = scorers.map((score) => {
+          return score.name;
+        });
+        const sampleScorer = sample.scores[scorer];
+        const scoreVal = sampleScorer.value;
+        if (typeof scoreVal === "object") {
+          const names = Object.keys(scoreVal);
+          if (
+            names.find((name) => {
+              return !scoreNames.includes(name);
+            })
+          ) {
+            // Since this dictionary contains keys which are not scores
+            // we just treat it like an opaque dictionary
+            return [
+              {
+                name: scorer,
+                rendered: () => {
+                  return scoreDescriptor.render(scoreVal);
+                },
+              },
+            ];
+          } else {
+            // Since this dictionary contains keys which are  scores
+            // we actually render the individual scores
+            const scores = names.map((name) => {
+              return {
+                name,
+                rendered: () => {
+                  return scoreDescriptor.render(scoreVal[name]);
+                },
+              };
+            });
+            return scores;
+          }
+        } else {
+          return [
+            {
+              name: scorer,
+              rendered: () => {
+                return scoreDescriptor.render(scoreVal);
+              },
+            },
+          ];
+        }
+      },
+    };
+  };
+
+  return {
+    scoreDescriptor,
+    epochs,
+    messageShape,
+    selectedScore: (sample) => {
+      return {
+        value: scoreValue(sample),
+        render: () => {
+          return scoreRendered(sample);
+        },
+      };
+    },
+    scorer: (sample, scorer) => {
+      return scorerDescriptor(sample, scorer);
+    },
+    selectedScorer: (sample) => {
+      return scorerDescriptor(sample, selectedScore?.scorer);
+    },
+  };
 };
 
 const scoreCategorizers = [
@@ -162,12 +337,11 @@ const scoreCategorizers = [
           scoreType: kScoreTypeObject,
           categories,
           render: (score) => {
-            if (score === null) {
+            if (score === null || score === undefined) {
               return "[null]";
             }
 
             const scores = [];
-
             const keys = Object.keys(score);
             keys.forEach((key, index) => {
               const value = score[key];
diff --git a/src/inspect_ai/_view/www/src/samples/SamplesTab.mjs b/src/inspect_ai/_view/www/src/samples/SamplesTab.mjs
index 414830f7c..8cb651036 100644
--- a/src/inspect_ai/_view/www/src/samples/SamplesTab.mjs
+++ b/src/inspect_ai/_view/www/src/samples/SamplesTab.mjs
@@ -16,6 +16,8 @@ export const SamplesTab = (props) => {
     sort,
     epoch,
     context,
+    selectedScore,
+    //setSelectedScore,
   } = props;
 
   const [selectedIndex, setSelectedIndex] = useState(0);
@@ -169,6 +171,7 @@ export const SamplesTab = (props) => {
         sampleDescriptor=${sampleDescriptor}
         selectedIndex=${selectedIndex}
         setSelectedIndex=${setSelectedIndex}
+        selectedScore=${selectedScore}
         nextSample=${nextSample}
         prevSample=${previousSample}
         showSample=${showSample}
diff --git a/src/inspect_ai/_view/www/src/samples/SamplesTools.mjs b/src/inspect_ai/_view/www/src/samples/SamplesTools.mjs
index 3dbacfd46..2a3eb5668 100644
--- a/src/inspect_ai/_view/www/src/samples/SamplesTools.mjs
+++ b/src/inspect_ai/_view/www/src/samples/SamplesTools.mjs
@@ -3,6 +3,7 @@ import { html } from "htm/preact";
 import { EpochFilter } from "./tools/EpochFilter.mjs";
 import { SortFilter } from "./tools/SortFilter.mjs";
 import { SampleFilter } from "./tools/SampleFilter.mjs";
+import { SelectScorer } from "./tools/SelectScorer.mjs";
 
 export const SampleTools = (props) => {
   const {
@@ -14,10 +15,24 @@ export const SampleTools = (props) => {
     setSort,
     epochs,
     sampleDescriptor,
+    score,
+    setScore,
+    scores,
   } = props;
 
   const hasEpochs = epochs > 1;
   const tools = [];
+
+  if (scores.length > 1) {
+    tools.push(
+      html`<${SelectScorer}
+        scores=${scores}
+        score=${score}
+        setScore=${setScore}
+      />`,
+    );
+  }
+
   if (hasEpochs) {
     tools.push(
       html`<${EpochFilter}
@@ -37,7 +52,12 @@ export const SampleTools = (props) => {
   );
 
   tools.push(
-    html`<${SortFilter} sort=${sort} setSort=${setSort} epochs=${hasEpochs} />`,
+    html`<${SortFilter}
+      sampleDescriptor=${sampleDescriptor}
+      sort=${sort}
+      setSort=${setSort}
+      epochs=${hasEpochs}
+    />`,
   );
 
   return tools;
diff --git a/src/inspect_ai/_view/www/src/samples/tools/SampleFilter.mjs b/src/inspect_ai/_view/www/src/samples/tools/SampleFilter.mjs
index ba8c6af25..cd42d7975 100644
--- a/src/inspect_ai/_view/www/src/samples/tools/SampleFilter.mjs
+++ b/src/inspect_ai/_view/www/src/samples/tools/SampleFilter.mjs
@@ -20,12 +20,13 @@ export const SampleFilter = ({ descriptor, filter, filterChanged }) => {
       filterChanged({
         value: val,
         filterFn: (sample, value) => {
-          if (typeof sample.score.value === "string") {
-            return sample.score.value.toLowerCase() === value?.toLowerCase();
-          } else if (typeof sample.score.value === "object") {
-            return JSON.stringify(sample.score.value) == value;
+          const score = descriptor.selectedScore(sample);
+          if (typeof score.value === "string") {
+            return score.value.toLowerCase() === value?.toLowerCase();
+          } else if (typeof score.value === "object") {
+            return JSON.stringify(score.value) == value;
           } else {
-            return sample.score.value === value;
+            return score.value === value;
           }
         },
       });
@@ -35,7 +36,7 @@ export const SampleFilter = ({ descriptor, filter, filterChanged }) => {
   const filterInput = (e) => {
     filterChanged({
       value: e.currentTarget.value,
-      filterFn: filterText,
+      filterFn: filterText(descriptor),
     });
   };
 
@@ -76,6 +77,7 @@ export const SampleFilter = ({ descriptor, filter, filterChanged }) => {
           class="form-control"
           value=${filter.value}
           placeholder="Filter Samples (score)"
+          style=${{ width: "150px" }}
           onInput=${filterInput}
         />
       `;
@@ -128,72 +130,75 @@ const SelectFilter = ({ value, options, filterFn }) => {
   `;
 };
 
-const filterText = (sample, value) => {
-  if (!value) {
-    return true;
-  } else {
-    if (isNumeric(value)) {
-      if (typeof sample.score.value === "number") {
-        return sample.score.value === Number(value);
-      } else {
-        return Number(sample.score.value) === Number(value);
-      }
+const filterText = (descriptor) => {
+  return (sample, value) => {
+    const score = descriptor.selectedScore(sample);
+    if (!value) {
+      return true;
     } else {
-      const filters = [
-        {
-          prefix: ">=",
-          fn: (score, val) => {
-            return score >= val;
+      if (isNumeric(value)) {
+        if (typeof score.value === "number") {
+          return score.value === Number(value);
+        } else {
+          return Number(score.value) === Number(value);
+        }
+      } else {
+        const filters = [
+          {
+            prefix: ">=",
+            fn: (score, val) => {
+              return score >= val;
+            },
           },
-        },
-        {
-          prefix: "<=",
-          fn: (score, val) => {
-            return score <= val;
+          {
+            prefix: "<=",
+            fn: (score, val) => {
+              return score <= val;
+            },
           },
-        },
-        {
-          prefix: ">",
-          fn: (score, val) => {
-            return score > val;
+          {
+            prefix: ">",
+            fn: (score, val) => {
+              return score > val;
+            },
           },
-        },
-        {
-          prefix: "<",
-          fn: (score, val) => {
-            return score < val;
+          {
+            prefix: "<",
+            fn: (score, val) => {
+              return score < val;
+            },
           },
-        },
-        {
-          prefix: "=",
-          fn: (score, val) => {
-            return score === val;
+          {
+            prefix: "=",
+            fn: (score, val) => {
+              return score === val;
+            },
           },
-        },
-        {
-          prefix: "!=",
-          fn: (score, val) => {
-            return score !== val;
+          {
+            prefix: "!=",
+            fn: (score, val) => {
+              return score !== val;
+            },
           },
-        },
-      ];
+        ];
 
-      for (const filter of filters) {
-        if (value?.startsWith(filter.prefix)) {
-          const val = value.slice(filter.prefix.length).trim();
-          if (!val) {
-            return true;
-          }
+        for (const filter of filters) {
+          if (value?.startsWith(filter.prefix)) {
+            const val = value.slice(filter.prefix.length).trim();
+            if (!val) {
+              return true;
+            }
 
-          const num = Number(val);
-          return filter.fn(sample.score.value, num);
+            const num = Number(val);
+            return filter.fn(score.value, num);
+          }
+        }
+        if (typeof score.value === "string") {
+          return score.value.toLowerCase() === value?.toLowerCase();
+        } else {
+          return score.value === value;
         }
-      }
-      if (typeof sample.score.value === "string") {
-        return sample.score.value.toLowerCase() === value?.toLowerCase();
-      } else {
-        return sample.score.value === value;
       }
     }
-  }
+  };
 };
diff --git a/src/inspect_ai/_view/www/src/samples/tools/SelectScorer.mjs b/src/inspect_ai/_view/www/src/samples/tools/SelectScorer.mjs
new file mode 100644
index 000000000..c6640305e
--- /dev/null
+++ b/src/inspect_ai/_view/www/src/samples/tools/SelectScorer.mjs
@@ -0,0 +1,117 @@
+import { html } from "htm/preact";
+
+export const SelectScorer = ({ scores, score, setScore }) => {
+  const scorers = scores.reduce((accum, scorer) => {
+    if (
+      !accum.find((sc) => {
+        return scorer.scorer === sc.scorer;
+      })
+    ) {
+      accum.push(scorer);
+    }
+    return accum;
+  }, []);
+
+  if (scorers.length === 1) {
+    // There is only a single scorer in play, just show the list of available scores
+    return html`
+      <div style=${{ display: "flex" }}>
+        <span class="select-scorer-label" style=${{ alignSelf: "center" }}
+          >Score:</span
+        >
+        <${ScoreSelector}
+          scores=${scores}
+          selectedIndex=${scoreIndex(score, scores)}
+          selectedIndexChanged=${(index) => {
+            setScore(scores[index]);
+          }}
+        />
+      </div>
+    `;
+  } else {
+    // selected scorer
+
+    const scorerScores = scores.filter((sc) => {
+      return sc.scorer === score.scorer;
+    });
+
+    const selectors = [
+      html`<${ScorerSelector}
+        scorers=${scorers}
+        selectedIndex=${scorerIndex(score, scorers)}
+        selectedIndexChanged=${(index) => {
+          setScore(scorers[index]);
+        }}
+      />`,
+    ];
+    if (scorerScores.length > 1) {
+      selectors.push(
+        html`<${ScoreSelector}
+          style=${{ marginLeft: "1em" }}
+          scores=${scorerScores}
+          selectedIndex=${scoreIndex(score, scorerScores)}
+          selectedIndexChanged=${(index) => {
+            setScore(scorerScores[index]);
+          }}
+        />`,
+      );
+    }
+
+    // There are multiple scorers, so show a scorer selector and a r
+    return html`
+      <div style=${{ display: "flex" }}>
+        <span class="select-scorer-label" style=${{ alignSelf: "center" }}
+          >Scorer:</span
+        >
+        ${selectors}
+      </div>
+    `;
+  }
+};
+
+const ScoreSelector = ({
+  scores,
+  selectedIndex,
+  selectedIndexChanged,
+  style,
+}) => {
+  return html`<select
+    class="form-select form-select-sm"
+    aria-label=".epoch-filter-label"
+    style=${{ fontSize: "0.7rem", ...style }}
+    value=${scores[selectedIndex].name}
+    onChange=${(e) => {
+      selectedIndexChanged(e.target.selectedIndex);
+    }}
+  >
+    ${scores.map((score) => {
+      return html`<option value="${score.name}">${score.name}</option>`;
+    })}
+  </select>`;
+};
+
+const ScorerSelector = ({ scorers, selectedIndex, selectedIndexChanged }) => {
+  return html`<select
+    class="form-select form-select-sm"
+    aria-label=".epoch-filter-label"
+    style=${{ fontSize: "0.7rem" }}
+    value=${scorers[selectedIndex].scorer}
+    onChange=${(e) => {
+      selectedIndexChanged(e.target.selectedIndex);
+    }}
+  >
+    ${scorers.map((scorer) => {
+      return html`<option value="${scorer.scorer}">${scorer.scorer}</option>`;
+    })}
+  </select>`;
+};
+
+const scoreIndex = (score, scores) =>
+  scores.findIndex((sc) => {
+    return sc.name === score.name && sc.scorer === score.scorer;
+  });
+
+const scorerIndex = (score, scores) =>
+  scores.findIndex((sc) => {
+    return sc.scorer === score.scorer;
+  });
diff --git a/src/inspect_ai/_view/www/src/samples/tools/SortFilter.mjs b/src/inspect_ai/_view/www/src/samples/tools/SortFilter.mjs
index e0a352efa..dbd067bdc 100644
--- a/src/inspect_ai/_view/www/src/samples/tools/SortFilter.mjs
+++ b/src/inspect_ai/_view/www/src/samples/tools/SortFilter.mjs
@@ -10,7 +10,7 @@ const kScoreDescVal = "score-desc";
 
 export const kDefaultSort = kSampleAscVal;
 
-export const SortFilter = ({ sort, setSort, epochs }) => {
+export const SortFilter = ({ sampleDescriptor, sort, setSort, epochs }) => {
   const options = [
     { label: "sample asc", val: kSampleAscVal },
     { label: "sample desc", val: kSampleDescVal },
@@ -25,14 +25,16 @@ export const SortFilter = ({ sort, setSort, epochs }) => {
       val: kEpochDescVal,
     });
   }
-  options.push({
-    label: "score asc",
-    val: kScoreAscVal,
-  });
-  options.push({
-    label: "score desc",
-    val: kScoreDescVal,
-  });
+  if (sampleDescriptor?.scoreDescriptor?.compare) {
+    options.push({
+      label: "score asc",
+      val: kScoreAscVal,
+    });
+    options.push({
+      label: "score desc",
+      val: kScoreDescVal,
+    });
+  }
   return html`
     <div style=${{ display: "flex" }}>
       <span
@@ -86,13 +88,13 @@ export const sort = (sort, samples, sampleDescriptor) => {
         return b.epoch - a.epoch;
       case kScoreAscVal:
         return sampleDescriptor.scoreDescriptor.compare(
-          a.score.value,
-          b.score.value,
+          sampleDescriptor.selectedScore(a).value,
+          sampleDescriptor.selectedScore(b).value,
         );
       case kScoreDescVal:
         return sampleDescriptor.scoreDescriptor.compare(
-          b.score.value,
-          a.score.value,
+          sampleDescriptor.selectedScore(b).value,
+          sampleDescriptor.selectedScore(a).value,
         );
     }
   });
diff --git a/src/inspect_ai/_view/www/src/sidebar/Sidebar.mjs b/src/inspect_ai/_view/www/src/sidebar/Sidebar.mjs
index b728e1a52..a729fb36e 100644
--- a/src/inspect_ai/_view/www/src/sidebar/Sidebar.mjs
+++ b/src/inspect_ai/_view/www/src/sidebar/Sidebar.mjs
@@ -53,10 +53,13 @@ export const Sidebar = ({
           <i class=${icons.close}></i>
         </button>
       </div>
-      <div style=${{ marginTop: "43px" }}>
-        <${ProgressBar} animating=${loading} style=${{ marginTop: "-2px" }} />
+      <div style=${{ marginTop: "41px", zIndex: 3 }}>
+        <${ProgressBar} animating=${loading} />
       </div>
-      <ul class="list-group" style=${{ flexGrow: 1, overflowY: "auto" }}>
+      <ul
+        class="list-group"
+        style=${{ flexGrow: 1, overflowY: "auto", marginTop: "-3px" }}
+      >
         ${logs.files.map((file, index) => {
           const active = index === selectedIndex ? " active" : "";
           const logHeader = logHeaders[file.name];
@@ -69,7 +72,15 @@ export const Sidebar = ({
 
           const model = logHeader?.eval?.model;
           const dataset = logHeader?.eval?.dataset;
-          const scorer = logHeader?.results?.scorer?.name;
+          const scorer = logHeader?.results?.scores
+            ?.map((scorer) => {
+              return scorer.name;
+            })
+            .join(",");
+          const scorerLabel =
+            Object.keys(logHeader?.results?.scores || {}).length === 1
+              ? "scorer"
+              : "scorers";
 
           const completed = logHeader?.stats?.completed_at;
           const time = completed ? new Date(completed) : undefined;
@@ -114,57 +125,7 @@ export const Sidebar = ({
                       </div>`
                     : ""}
                 </div>
-                ${logHeader?.results?.metrics
-                  ? html`<div
-                      style=${{
-                        display: "flex",
-                        flexDirection: "row",
-                        flexWrap: "wrap",
-                        justifyContent: "flex-end",
-                      }}
-                    >
-                      ${Object.keys(logHeader?.results.metrics).map(
-                        (metric) => {
-                          return html`
-                            <div
-                              style=${{
-                                display: "flex",
-                                flexDirection: "column",
-                                alignItems: "center",
-                                marginLeft: "1em",
-                              }}
-                            >
-                              <div style=${{ fontWeight: 300 }}>
-                                ${logHeader?.results.metrics[metric].name}
-                              </div>
-                              <div
-                                style=${{ fontWeight: 600, fontSize: "1.5em" }}
-                              >
-                                ${formatPrettyDecimal(
-                                  logHeader?.results.metrics[metric].value,
-                                )}
-                              </div>
-                            </div>
-                          `;
-                        },
-                      )}
-                    </div>`
-                  : logHeader?.status === "error"
-                    ? html`<div style=${{ color: "var(--bs-danger)" }}>
-                        Eval Error
-                      </div>`
-                    : logHeader?.status === "cancelled"
-                      ? html`<div style=${{ color: "var(--bs-secondary)" }}>
-                          Cancelled
-                        </div>`
-                      : logHeader?.status === "started"
-                        ? html`<div
-                            class="spinner-border spinner-border-sm"
-                            role="status"
-                          >
-                            <span class="visually-hidden">Loading...</span>
-                          </div>`
-                        : ""}
+                <${EvalStatus} logHeader=${logHeader} />
               </div>
               <div style=${{ marginTop: "0.4em" }}>
                 <small class="mb-1 text-muted">
@@ -186,7 +147,7 @@ export const Sidebar = ({
                     }}
                   >
                     <span>dataset: ${dataset.name || "(samples)"}</span
-                    ><span>scorer: ${scorer}</span>
+                    ><span>${scorerLabel}: ${scorer}</span>
                   </div>`
                 : ""}
             </li>
@@ -212,6 +173,129 @@ const prettyDir = (path) => {
   }
 };
 
+const EvalStatus = ({ logHeader }) => {
+  switch (logHeader.status) {
+    case "error":
+      return html`<${StatusError} message="Error" />`;
+
+    case "cancelled":
+      return html`<${StatusCancelled} message="Cancelled" />`;
+
+    case "started":
+      return html`<${StatusRunning} message="Running" />`;
+
+    default:
+      if (logHeader?.results?.scores && logHeader.results.scores.length > 0) {
+        if (logHeader.results.scores.length === 1) {
+          return html`<${SidebarScore}
+            scorer=${logHeader.results.scores[0]}
+          />`;
+        } else {
+          return html`<${SidebarScores} scores=${logHeader.results.scores} />`;
+        }
+      } else {
+        return "";
+      }
+  }
+};
+
+const SidebarScore = ({ scorer }) => {
+  return html`<div
+    style=${{
+      display: "flex",
+      flexDirection: "row",
+      flexWrap: "wrap",
+      justifyContent: "flex-end",
+    }}
+  >
+    ${Object.keys(scorer.metrics).map((metric) => {
+      return html`
+        <div
+          style=${{
+            display: "flex",
+            flexDirection: "column",
+            alignItems: "center",
+            marginLeft: "1em",
+          }}
+        >
+          <div style=${{ fontWeight: 300 }}>${scorer.metrics[metric].name}</div>
+          <div style=${{ fontWeight: 600, fontSize: "1.5em" }}>
+            ${formatPrettyDecimal(scorer.metrics[metric].value)}
+          </div>
+        </div>
+      `;
+    })}
+  </div>`;
+};
+
+const SidebarScores = ({ scores }) => {
+  return html`<div
+    style=${{
+      display: "flex",
+      flexDirection: "row",
+      flexWrap: "wrap",
+      justifyContent: "flex-end",
+      rowGap: "1em",
+    }}
+  >
+    ${scores.map((score) => {
+      const name = score.name;
+      return html`
+        <div
+          style=${{
+            display: "flex",
+            flexDirection: "column",
+            alignItems: "center",
+            marginLeft: "1em",
+          }}
+        >
+          <div
+            style=${{
+              fontSize: "0.6rem",
+              width: "100%",
+              fontWeight: 300,
+              borderBottom: "solid var(--bs-border-color) 1px",
+              textTransform: "uppercase",
+            }}
+          >
+            ${name}
+          </div>
+          <div
+            style=${{
+              fontSize: "0.7rem",
+              display: "grid",
+              gridTemplateColumns: "max-content max-content",
+              gridGap: "0 0.3rem",
+            }}
+          >
+            ${Object.keys(score.metrics).map((key) => {
+              const metric = score.metrics[key];
+              return html` <div>${metric.name}</div>
+                <div style=${{ fontWeight: "600" }}>
+                  ${formatPrettyDecimal(metric.value)}
+                </div>`;
+            })}
+          </div>
+        </div>
+      `;
+    })}
+  </div>`;
+};
+
+const StatusCancelled = ({ message }) => {
+  return html`<div style=${{ color: "var(--bs-secondary)" }}>${message}</div>`;
+};
+
+const StatusRunning = ({ message }) => {
+  return html`<div class="spinner-border spinner-border-sm" role="status">
+    <span class="visually-hidden">${message}</span>
+  </div>`;
+};
+
+const StatusError = ({ message }) => {
+  return html`<div style=${{ color: "var(--bs-danger)" }}>${message}</div>`;
+};
+
 const LogDirectoryTitle = ({ log_dir, offcanvas }) => {
   if (log_dir) {
     const displayDir = prettyDir(log_dir);
diff --git a/src/inspect_ai/_view/www/src/title/TitleBlock.mjs b/src/inspect_ai/_view/www/src/title/TitleBlock.mjs
index 97d481166..1cfc02d3f 100644
--- a/src/inspect_ai/_view/www/src/title/TitleBlock.mjs
+++ b/src/inspect_ai/_view/www/src/title/TitleBlock.mjs
@@ -35,11 +35,12 @@ export const TitleBlock = ({ log, status }) => {
 `,
   });
 
+  const label = log?.results?.scores.length > 1 ? "Scorers" : "Scorer";
   values.push({
     size: "auto",
-    value: html`<${LabeledValue} label="Scorer" style=${staticColStyle}>
+    value: html`<${LabeledValue} label="${label}" style=${staticColStyle}>
     <${ScorerSummary} 
-    scorer=${log?.results?.scorer} />
+      scorers=${log?.results?.scores} />
   </${LabeledValue}>`,
   });
 
@@ -90,13 +91,19 @@ const DatasetSummary = ({ dataset, samples, epochs, style }) => {
   `;
 };
 
-const ScorerSummary = ({ scorer }) => {
-  if (!scorer) {
+const ScorerSummary = ({ scorers }) => {
+  if (!scorers) {
     return "";
   }
 
   const summary = [];
-  summary.push(scorer.name);
+  summary.push(
+    scorers
+      .map((scorer) => {
+        return scorer.name;
+      })
+      .join(", "),
+  );
   return summary;
 };
 
diff --git a/src/inspect_ai/_view/www/src/utils/Format.mjs b/src/inspect_ai/_view/www/src/utils/Format.mjs
index 157d50bd7..ca30897d5 100644
--- a/src/inspect_ai/_view/www/src/utils/Format.mjs
+++ b/src/inspect_ai/_view/www/src/utils/Format.mjs
@@ -36,24 +36,6 @@ export const shortenCompletion = (completion) => {
   return shortened || completion;
 };
 
-export const answerForSample = (sample) => {
-  if (sample) {
-    if (sample.score?.answer) {
-      return sample.score.answer;
-    } else if (sample.output.choices && sample.output.choices.length > 0) {
-      const content = sample.output.choices[0].message.content;
-      if (typeof content === "string") {
-        return content;
-      } else {
-        // TODO: Support image completions.
-        return content.length > 0 ? content[0].text : "";
-      }
-    }
-  } else {
-    return undefined;
-  }
-};
-
 // Gets a string for a sample input
 export const inputString = (input) => {
   if (typeof input === "string") {
diff --git a/src/inspect_ai/_view/www/src/workspace/WorkSpace.mjs b/src/inspect_ai/_view/www/src/workspace/WorkSpace.mjs
index 3e88151c4..05fc2a21d 100644
--- a/src/inspect_ai/_view/www/src/workspace/WorkSpace.mjs
+++ b/src/inspect_ai/_view/www/src/workspace/WorkSpace.mjs
@@ -36,23 +36,21 @@ export const WorkSpace = (props) => {
   const divRef = useRef();
   const codeRef = useRef();
 
+  // alias the log for the workspace
   const workspaceLog = props.log;
+
+  // State tracking for the view
   const [currentTaskId, setCurrentTaskId] = useState(
     workspaceLog?.contents?.eval?.run_id,
   );
-
-  // State tracking for the view
-  const [state, setState] = useState({
-    logFiltered: undefined,
-    viewState: {
-      selectedTab: kEvalTabId,
-      openSamples: [],
-      filter: {},
-      epoch: "all",
-      sort: kDefaultSort,
-      renderedCode: false,
-    },
-  });
+  const [selectedTab, setSelectedTab] = useState(kEvalTabId);
+  const [scores, setScores] = useState([]);
+  const [score, setScore] = useState(undefined);
+  const [samplesDesc, setSamplesDesc] = useState(undefined);
+  const [filter, setFilter] = useState({});
+  const [epoch, setEpoch] = useState("all");
+  const [sort, setSort] = useState(kDefaultSort);
+  const [renderedCode, setRenderedCode] = useState(false);
 
   // Context is shared with most/all components and
   // allows for global information to pass between components
@@ -63,12 +61,68 @@ export const WorkSpace = (props) => {
     },
   };
 
-  const sampleDescriptor = useMemo(() => {
-    return samplesDescriptor(
+  const clearSampleTools = useCallback(() => {
+    setEpoch("all");
+    setFilter({});
+    setSort(kDefaultSort);
+  }, [setEpoch, setFilter, setSort]);
+
+  // Display the log
+  useEffect(() => {
+    if (
+      workspaceLog.contents &&
+      workspaceLog.contents.eval?.run_id !== currentTaskId
+    ) {
+      const defaultTab =
+        workspaceLog.contents?.status !== "error" ? kEvalTabId : kInfoTabId;
+      setSelectedTab(defaultTab);
+      if (divRef.current) {
+        divRef.current.scrollTop = 0;
+      }
+    }
+  }, [workspaceLog, divRef, currentTaskId, setSelectedTab]);
+
+  useEffect(() => {
+    // Select the default scorer to use
+    const scorer = workspaceLog?.contents?.results?.scores[0]
+      ? {
+          name: workspaceLog.contents.results.scores[0].name,
+          scorer: workspaceLog.contents.results.scores[0].scorer,
+        }
+      : undefined;
+    const scorers = (workspaceLog.contents?.results?.scores || []).map(
+      (score) => {
+        return {
+          name: score.name,
+          scorer: score.scorer,
+        };
+      },
+    );
+
+    // Reset state
+    setScores(scorers);
+    setScore(scorer);
+    clearSampleTools();
+    setRenderedCode(false);
+  }, [workspaceLog, setScores, setScore, setEpoch, setFilter, setRenderedCode]);
+
+  useEffect(() => {
+    clearSampleTools();
+  }, [score]);
+
+  useEffect(() => {
+    const sampleDescriptor = samplesDescriptor(
+      score,
+      scores,
       workspaceLog.contents?.samples,
       workspaceLog.contents?.eval?.config?.epochs || 1,
       context,
     );
+    setSamplesDesc(sampleDescriptor);
+  }, [workspaceLog, score, scores, setSamplesDesc]);
+
+  useEffect(() => {
+    setCurrentTaskId(workspaceLog.contents?.eval?.run_id);
   }, [workspaceLog]);
 
   // Tabs that are available within the app
@@ -89,28 +143,33 @@ export const WorkSpace = (props) => {
           return html` <${SamplesTab}
             task=${workspaceLog.contents?.eval?.task}
             model=${workspaceLog.contents?.eval?.model}
+            selectedScore=${score}
+            setSelectedScore=${setScore}
             samples=${workspaceLog.contents?.samples}
-            sampleDescriptor=${sampleDescriptor}
-            filter=${state.viewState.filter}
-            sort=${state.viewState.sort}
-            epoch=${state.viewState.epoch}
+            sampleDescriptor=${samplesDesc}
+            filter=${filter}
+            sort=${sort}
+            epoch=${epoch}
             context=${context}
           />`;
         },
-        tools: (state) => {
+        tools: () => {
           // Don't show tools if there is a sample sample
           if (workspaceLog.contents?.samples?.length <= 1) {
             return "";
           }
           return html`<${SampleTools}
-            epoch=${state.viewState.epoch}
+            epoch=${epoch}
             epochs=${workspaceLog.contents?.eval?.config?.epochs}
             setEpoch=${setEpoch}
-            filter=${state.viewState.filter}
-            filterChanged=${filterChanged}
-            sort=${state.viewState.sort}
+            filter=${filter}
+            filterChanged=${setFilter}
+            sort=${sort}
             setSort=${setSort}
-            sampleDescriptor=${sampleDescriptor}
+            score=${score}
+            setScore=${setScore}
+            scores=${scores}
+            sampleDescriptor=${samplesDesc}
           />`;
         },
       };
@@ -190,7 +249,7 @@ export const WorkSpace = (props) => {
             />`,
           );
         } else {
-          if (codeRef.current && !state.viewState.renderedCode) {
+          if (codeRef.current && !renderedCode) {
             if (workspaceLog.raw.length < kPrismRenderMaxSize) {
               codeRef.current.innerHTML = Prism.highlight(
                 workspaceLog.raw,
@@ -203,9 +262,7 @@ export const WorkSpace = (props) => {
               codeRef.current.appendChild(textNode);
             }
 
-            const viewState = state.viewState;
-            viewState.renderedCode = true;
-            setState({ viewState });
+            setRenderedCode(true);
           }
           renderedContent.push(
             html`<pre>
@@ -248,40 +305,18 @@ export const WorkSpace = (props) => {
     };
 
     return resolvedTabs;
-  }, [state, workspaceLog]);
-
-  const setSelectedTab = (currentState, selectedTab) => {
-    const viewState = currentState.viewState;
-    viewState.selectedTab = selectedTab;
-    setState({ viewState });
-  };
-
-  const filterChanged = useCallback(
-    (filter) => {
-      const viewState = state.viewState;
-      viewState.filter = filter;
-      setState({ viewState });
-    },
-    [state, setState],
-  );
-
-  const setEpoch = useCallback(
-    (epoch) => {
-      const viewState = state.viewState;
-      viewState.epoch = epoch;
-      setState({ viewState });
-    },
-    [state],
-  );
-
-  const setSort = useCallback(
-    (sort) => {
-      const viewState = state.viewState;
-      viewState.sort = sort;
-      setState({ viewState });
-    },
-    [state],
-  );
+  }, [
+    samplesDesc,
+    workspaceLog,
+    filter,
+    setFilter,
+    epoch,
+    setEpoch,
+    sort,
+    setSort,
+    renderedCode,
+    setRenderedCode,
+  ]);
 
   const copyFeedback = useCallback(
     (e) => {
@@ -301,36 +336,9 @@ export const WorkSpace = (props) => {
         }, 1250);
       }
     },
-    [state],
+    [renderedCode],
   );
 
-  // Display the log
-  useEffect(() => {
-    if (workspaceLog.contents && workspaceLog.eval?.run_id !== currentTaskId) {
-      const defaultTab =
-        workspaceLog.contents?.status !== "error" ? kEvalTabId : kInfoTabId;
-      setSelectedTab(state, defaultTab);
-      if (divRef.current) {
-        divRef.current.scrollTop = 0;
-      }
-    }
-
-    // Reset state
-    const newState = {
-      openSamples: [],
-      filter: {},
-      epoch: "all",
-      sort: kDefaultSort,
-      renderedCode: false,
-    };
-
-    setState({ viewState: { ...state.viewState, ...newState } });
-  }, [workspaceLog, divRef, currentTaskId]);
-
-  useEffect(() => {
-    setCurrentTaskId(workspaceLog.contents?.eval?.run_id);
-  }, [workspaceLog]);
-
   // Compute the tools for this tab
   const tabTools = Object.keys(tabs)
     .map((key) => {
@@ -338,32 +346,27 @@ export const WorkSpace = (props) => {
       return tab;
     })
     .filter((tab) => {
-      return tab.id === state.viewState.selectedTab;
+      return tab.id === selectedTab;
     })
     .map((tab) => {
       if (tab.tools) {
-        const tools = tab.tools(state);
+        const tools = tab.tools();
         return tools;
       } else {
         return "";
       }
     });
 
-  const selectTab = (event) => {
-    const id = event.currentTarget.id;
-    setSelectedTab(state, id);
-  };
-
   return html`<${WorkspaceDisplay}
     divRef=${divRef}
     tabs=${tabs}
     tabTools=${tabTools}
     log=${workspaceLog}
-    selectedTab=${state.viewState.selectedTab}
+    selectedTab=${selectedTab}
     fullScreen=${props.fullScreen}
     offcanvas=${props.offcanvas}
     context=${context}
-    selectTab=${selectTab}
+    setSelectedTab=${setSelectedTab}
     afterBodyElements=${afterBodyElements}
   />`;
 };
@@ -373,9 +376,7 @@ const WorkspaceDisplay = ({
   selectedTab,
   tabs,
   tabTools,
-  selectTab,
-  fullScreen,
-  offcanvas,
+  setSelectedTab,
   divRef,
   context,
   afterBodyElements,
@@ -383,11 +384,9 @@ const WorkspaceDisplay = ({
   if (log.contents === undefined) {
     return html`<${EmptyPanel} />`;
   } else {
-    const fullScreenClz = fullScreen ? " full-screen" : "";
-    const offcanvasClz = offcanvas ? " off-canvas" : "";
-
-    return html`<div ref=${divRef} class="workspace${fullScreenClz}${offcanvasClz}" style=${{
+    return html`<div ref=${divRef} class="workspace" style=${{
       paddingTop: "0rem",
+      overflowY: "hidden",
     }}>
             <${TitleBlock}
               created=${log.contents?.eval.created}
@@ -426,7 +425,10 @@ const WorkspaceDisplay = ({
                 return html`<${TabPanel}
                 id=${tab.id}
                 title="${tab.label}"
-                onSelected="${selectTab}"
+                onSelected=${(e) => {
+                  const id = e.currentTarget.id;
+                  setSelectedTab(id);
+                }}
                 selected=${selectedTab === tab.id}
                 scrollable=${!!tab.scrollable}>
                   ${tab.content()}
diff --git a/src/inspect_ai/log/__init__.py b/src/inspect_ai/log/__init__.py
index 14fb46f13..6763178fd 100644
--- a/src/inspect_ai/log/__init__.py
+++ b/src/inspect_ai/log/__init__.py
@@ -15,7 +15,7 @@
     EvalResults,
     EvalRevision,
     EvalSample,
-    EvalScorer,
+    EvalScore,
     EvalSpec,
     EvalStats,
     LoggingLevel,
@@ -34,7 +34,7 @@
     "EvalResults",
     "EvalRevision",
     "EvalSample",
-    "EvalScorer",
+    "EvalScore",
     "EvalSpec",
     "EvalStats",
     "EvalLogInfo",
diff --git a/src/inspect_ai/log/_file.py b/src/inspect_ai/log/_file.py
index 2e9803097..e77b8a2b3 100644
--- a/src/inspect_ai/log/_file.py
+++ b/src/inspect_ai/log/_file.py
@@ -27,6 +27,8 @@
     Recorder,
 )
 
+LOG_SCHEMA_VERSION = 2
+
 
 class EvalLogInfo(FileInfo):
     task: str
@@ -120,7 +122,7 @@ def read_eval_log(log_file: str | FileInfo, header_only: bool = False) -> EvalLo
 
     # verify we know about this version of the log file format
     def validate_version(ver: int) -> None:
-        if ver > 1:
+        if ver > LOG_SCHEMA_VERSION:
             raise ValueError(f"Unable to read version {ver} of log format.")
 
     # header-only uses json-stream
@@ -136,8 +138,10 @@ def read_field(field: str) -> Any:
                         return None
 
                 # fail for unknown version
-                version = read_field("version")
-                validate_version(version)
+                validate_version(read_field("version"))
+
+                # set the version to the schema version we'll be returning
+                version = LOG_SCHEMA_VERSION
 
                 results = read_field("results")
                 error = read_field("error")
@@ -171,6 +175,9 @@ def read_field(field: str) -> Any:
         # fail for unknown version
         validate_version(log.version)
 
+        # set the version to the schema version we'll be returning
+        log.version = LOG_SCHEMA_VERSION
+
         # prune if header_only
         if header_only:
             log.samples = None
diff --git a/src/inspect_ai/log/_log.py b/src/inspect_ai/log/_log.py
index 4e5713984..a7604a2d2 100644
--- a/src/inspect_ai/log/_log.py
+++ b/src/inspect_ai/log/_log.py
@@ -1,5 +1,6 @@
 import abc
 import asyncio
+import logging
 import os
 import sys
 import traceback
@@ -9,7 +10,7 @@
 
 import click
 import tenacity
-from pydantic import BaseModel, ConfigDict, Field
+from pydantic import BaseModel, ConfigDict, Field, model_validator
 from rich.console import Console, RenderableType
 from rich.traceback import Traceback
 
@@ -23,6 +24,8 @@
 )
 from inspect_ai.scorer import Score
 
+SCORER_PLACEHOLDER = "88F74D2C"
+
 
 class EvalConfig(BaseModel):
     limit: int | tuple[int, int] | None = Field(default=None)
@@ -75,12 +78,41 @@ class EvalSample(BaseModel):
     output: ModelOutput
     """Model output from sample."""
 
-    score: Score | None = Field(default=None)
-    """Score for sample."""
+    @property
+    def score(self) -> Score | None:
+        """Score for sample (deprecated)."""
+        logging.warning(
+            "The 'score' field is deprecated. Access sample scores through 'scores' instead."
+        )
+        return list(self.scores.values())[0] if self.scores else None
+
+    scores: dict[str, Score] | None = Field(default=None)
+    """Scores for sample."""
 
     metadata: dict[str, Any]
     """Additional sample metadata."""
 
+    @model_validator(mode="before")
+    @classmethod
+    def convert_score_to_scores(
+        cls: Type["EvalSample"], values: dict[str, Any]
+    ) -> dict[str, Any]:
+        if "score" in values:
+            # There cannot be a scorers property too
+            if "scores" in values:
+                raise TypeError(
+                    "Unexpected value `scores` present when `score` has already been specified."
+                )
+
+            # Convert the scorer to the new schema
+            score = values["score"]
+            values["scores"] = {SCORER_PLACEHOLDER: score}
+
+            # Get rid of the 'scorer' property
+            del values["score"]
+
+        return values
+
 
 class EvalPlanStep(BaseModel):
     solver: str
@@ -90,17 +122,6 @@ class EvalPlanStep(BaseModel):
     """Parameters used to instantiate solver."""
 
 
-class EvalScorer(BaseModel):
-    name: str
-    """Scorer name."""
-
-    params: dict[str, Any] = Field(default={})
-    """Parameters specified when creating scorer."""
-
-    metadata: dict[str, Any] | None = Field(default=None)
-    """Additional scorer metadata."""
-
-
 class EvalPlan(BaseModel):
     name: str = Field(default="plan")
     """Plan name."""
@@ -129,16 +150,74 @@ class EvalMetric(BaseModel):
     """Additional metadata associated with metric."""
 
 
+class EvalScore(BaseModel):
+    name: str
+    """Score name."""
+
+    scorer: str
+    """Scorer name."""
+
+    params: dict[str, Any] = Field(default={})
+    """Parameters specified when creating scorer."""
+
+    metrics: dict[str, EvalMetric] = Field(default=[])
+    """Metrics computed for this scorer."""
+
+    metadata: dict[str, Any] | None = Field(default=None)
+    """Additional scorer metadata."""
+
+
 class EvalResults(BaseModel):
-    scorer: EvalScorer | None = Field(default=None)
-    """Scorer used to compute results"""
+    @property
+    def scorer(self) -> EvalScore | None:
+        """Scorer used to compute results (deprecated)."""
+        logging.warning(
+            "The 'scorer' field is deprecated. Use 'scorers' instead.",
+        )
+        return self.scores[0] if self.scores else None
+
+    @property
+    def metrics(self) -> dict[str, EvalMetric]:
+        """Metrics computed (deprecated)."""
+        logging.warning(
+            "The 'metrics' field is deprecated. Access metrics through 'scorers' instead."
+        )
+        return self.scores[0].metrics if self.scores else {}
 
-    metrics: dict[str, EvalMetric] = Field(default={})
-    """Metrics computed."""
+    scores: list[EvalScore] = Field(default=[])
+    """Scorers used to compute results"""
 
     metadata: dict[str, Any] | None = Field(default=None)
     """Additional results metadata."""
 
+    @model_validator(mode="before")
+    @classmethod
+    def convert_scorer_to_scorers(
+        cls: Type["EvalResults"], values: dict[str, Any]
+    ) -> dict[str, Any]:
+        if "scorer" in values:
+            # There cannot be a scorers property too
+            if "scores" in values:
+                raise TypeError(
+                    "Unexpected value `scores` present when `scorer` has already been specified."
+                )
+
+            # Gather metrics
+            if "metrics" in values:
+                metrics = values["metrics"]
+                del values["metrics"]
+            # Convert the scorer to the new schema
+            score = values["scorer"]
+            if metrics:
+                score["metrics"] = metrics
+            score["scorer"] = score["name"]
+            values["scores"] = [score]
+
+            # Get rid of the 'scorer' property
+            del values["scorer"]
+
+        return values
+
 
 class EvalDataset(BaseModel):
     name: str | None = Field(default=None)
@@ -316,7 +395,7 @@ def from_log_record(record: LogRecord) -> "LoggingMessage":
 
 
 class EvalLog(BaseModel):
-    version: int = Field(default=1)
+    version: int = Field(default=2)
     """Eval log file format version."""
 
     status: Literal["started", "success", "cancelled", "error"] = Field(
@@ -345,6 +424,17 @@ class EvalLog(BaseModel):
     logging: list[LoggingMessage] = Field(default=[])
     """Logging message captured during eval."""
 
+    @model_validator(mode="after")
+    def populate_scorer_name_for_samples(self) -> "EvalLog":
+        if self.samples and self.results and self.results.scores:
+            scorer_name = self.results.scores[0].name
+            for sample in self.samples:
+                if sample.scores and SCORER_PLACEHOLDER in sample.scores:
+                    sample.scores[scorer_name] = sample.scores[SCORER_PLACEHOLDER]
+                    del sample.scores[SCORER_PLACEHOLDER]
+
+        return self
+
 
 LogEvent = Literal["plan", "sample", "score", "results", "scorer", "logging"]
 
diff --git a/src/inspect_ai/scorer/_metric.py b/src/inspect_ai/scorer/_metric.py
index 9f14b370b..1baf31217 100644
--- a/src/inspect_ai/scorer/_metric.py
+++ b/src/inspect_ai/scorer/_metric.py
@@ -153,7 +153,7 @@ class Metric(Protocol):
         Metric value
     """
 
-    def __call__(self, scores: list[Score]) -> int | float: ...
+    def __call__(self, scores: list[Score]) -> Value: ...
 
 
 MetricType = TypeVar("MetricType", Callable[..., Metric], type[Metric])
diff --git a/src/inspect_ai/scorer/_scorer.py b/src/inspect_ai/scorer/_scorer.py
index 4e894c463..b77bfb6eb 100644
--- a/src/inspect_ai/scorer/_scorer.py
+++ b/src/inspect_ai/scorer/_scorer.py
@@ -87,12 +87,14 @@ def scorer_create(name: str, **kwargs: Any) -> Scorer:
 
 
 def scorer(
-    metrics: list[Metric], name: str | None = None, **metadata: Any
+    metrics: list[Metric] | dict[str, list[Metric]],
+    name: str | None = None,
+    **metadata: Any,
 ) -> Callable[[Callable[..., Scorer]], Callable[..., Scorer]]:
     r"""Decorator for registering scorers.
 
     Args:
-        metrics (list[Metric]): One or more metrics to calculate
+        metrics (list[Metric] | dict[str, list[Metric]]): One or more metrics to calculate
             over the scores.
         name (str | None):
             Optional name for scorer. If the decorator has no name
@@ -144,8 +146,12 @@ def scorer_wrapper(*args: Any, **kwargs: Any) -> Scorer:
     return wrapper
 
 
-def scorer_metrics(scorer: Scorer) -> list[Metric]:
-    return cast(list[Metric], registry_info(scorer).metadata[SCORER_METRICS])
+def scorer_metrics(scorer: Scorer) -> list[Metric] | dict[str, list[Metric]]:
+    metrics_raw = registry_info(scorer).metadata[SCORER_METRICS]
+    if isinstance(metrics_raw, dict):
+        return cast(dict[str, list[Metric]], metrics_raw)
+    else:
+        return cast(list[Metric], metrics_raw)
 
 
 SCORER_METRICS = "metrics"
diff --git a/tests/log/test_eval_log.py b/tests/log/test_eval_log.py
index daf7a9fc6..5e640e63b 100644
--- a/tests/log/test_eval_log.py
+++ b/tests/log/test_eval_log.py
@@ -59,7 +59,7 @@ def test_fail_invalid():
 
 
 def test_fail_version():
-    check_log_raises(log_path("log_version_2"))
+    check_log_raises(log_path("log_version_3"))
 
 
 def check_log_raises(log_file):
diff --git a/tests/log/test_eval_log/log_version_2.txt b/tests/log/test_eval_log/log_version_3.txt
similarity index 75%
rename from tests/log/test_eval_log/log_version_2.txt
rename to tests/log/test_eval_log/log_version_3.txt
index 63e92fb01..470e79b73 100644
--- a/tests/log/test_eval_log/log_version_2.txt
+++ b/tests/log/test_eval_log/log_version_3.txt
@@ -1,5 +1,5 @@
 {
-    "version": 2,
+    "version": 3,
     "status": "success",
     "eval": {
         "task": "wikipedia",
@@ -31,22 +31,22 @@
         "config": {}
     },
     "results": {
-        "scorer": {
+        "scorers": [{
             "name": "model_graded_fact",
             "params": {}
-        },
-        "metrics": {
-            "accuracy": {
-                "name": "accuracy",
-                "value": 1,
-                "options": {}
-            },
-            "bootstrap_std": {
-                "name": "bootstrap_std",
-                "value": 0.0,
-                "options": {}
+            "metrics": {
+                "accuracy": {
+                    "name": "accuracy",
+                    "value": 1,
+                    "options": {}
+                },
+                "bootstrap_std": {
+                    "name": "bootstrap_std",
+                    "value": 0.0,
+                    "options": {}
+                }
             }
-        }
+        }]
     },
     "stats": {
         "started_at": "2024-05-05T07:59:35",
diff --git a/tests/scorer/test_metric.py b/tests/scorer/test_metric.py
index 7a4d2ef33..0d1a6a39e 100644
--- a/tests/scorer/test_metric.py
+++ b/tests/scorer/test_metric.py
@@ -96,7 +96,7 @@ def check_log(log):
     check_log(log)
 
     # eval log w/ different scorer (that still uses accuracy)
-    log = score(log, scorer=includes())
+    log = score(log, scorers=[includes()])
     check_log(log)
 
 
diff --git a/tests/scorer/test_multiscorer.py b/tests/scorer/test_multiscorer.py
new file mode 100644
index 000000000..d2a877b4d
--- /dev/null
+++ b/tests/scorer/test_multiscorer.py
@@ -0,0 +1,103 @@
+import random
+
+from inspect_ai import Task, eval
+from inspect_ai.dataset import Sample
+from inspect_ai.scorer import Score, Target, bootstrap_std, mean, scorer
+from inspect_ai.solver import TaskState
+
+
+@scorer(metrics=[mean(), bootstrap_std()])
+def rand_score():
+    async def score(state: TaskState, target: Target):
+        answer = state.output.completion
+        return Score(value=random.randint(1, 100), answer=answer)
+
+    return score
+
+
+@scorer(metrics=[mean(), bootstrap_std()])
+def another_rand_score():
+    async def score(state: TaskState, target: Target):
+        answer = state.output.completion
+        return Score(value=random.randint(1, 100), answer=answer)
+
+    return score
+
+
+@scorer(
+    metrics={"a_count": [mean(), bootstrap_std()], "e_count": [mean(), bootstrap_std()]}
+)
+def letter_count():
+    async def score(state: TaskState, target: Target):
+        answer = state.output.completion
+        a_count = answer.count("a")
+        e_count = answer.count("e")
+        return Score(value={"a_count": a_count, "e_count": e_count}, answer=answer)
+
+    return score
+
+
+def check_log(log, scorers, metrics):
+    # core checks
+    assert log.results
+    assert log.results.scores
+    assert len(log.results.scores) == len(scorers)
+
+    scorer_names = [scorer.name for scorer in log.results.scores]
+    assert all(scorer in scorer_names for scorer in scorers)
+    assert all(
+        all(metric in scorer.metrics for metric in metrics)
+        for scorer in log.results.scores
+    )
+
+    # test deprecated fields for now
+    assert log.results.scorer is not None
+    assert log.results.metrics is not None
+
+
+# test a single scorer
+def test_single_scorer() -> None:
+    task = Task(
+        dataset=[Sample(input="What is 1 + 1?", target=["2", "2.0", "Two"])],
+        scorer=rand_score(),
+    )
+
+    # normal eval
+    log = eval(tasks=task, model="mockllm/model")[0]
+    check_log(log, ["rand_score"], ["mean", "bootstrap_std"])
+
+
+# test two scorers
+def test_multi_scorer() -> None:
+    task = Task(
+        dataset=[Sample(input="What is 1 + 1?", target=["2", "2.0", "Two"])],
+        scorer=[rand_score(), another_rand_score()],
+    )
+
+    # normal eval
+    log = eval(tasks=task, model="mockllm/model")[0]
+    check_log(log, ["rand_score", "another_rand_score"], ["mean", "bootstrap_std"])
+
+
+# test dictionary scorer
+def test_dict_scorer() -> None:
+    task = Task(
+        dataset=[Sample(input="What is 1 + 1?", target=["2", "2.0", "Two"])],
+        scorer=letter_count(),
+    )
+
+    # normal eval
+    log = eval(tasks=task, model="mockllm/model")[0]
+    check_log(log, ["a_count", "e_count"], ["mean", "bootstrap_std"])
+
+
+# test blend of dictionary and simple scorers
+def test_blend_scorer() -> None:
+    task = Task(
+        dataset=[Sample(input="What is 1 + 1?", target=["2", "2.0", "Two"])],
+        scorer=[letter_count(), rand_score()],
+    )
+
+    # normal eval
+    log = eval(tasks=task, model="mockllm/model")[0]
+    check_log(log, ["a_count", "e_count", "rand_score"], ["mean", "bootstrap_std"])
diff --git a/tools/vscode/assets/www/view/view-overrides.css b/tools/vscode/assets/www/view/view-overrides.css
index aba01b072..ebbff341c 100644
--- a/tools/vscode/assets/www/view/view-overrides.css
+++ b/tools/vscode/assets/www/view/view-overrides.css
@@ -46,6 +46,5 @@ body[class^="vscode-"] code:not(.sourceCode) {
    to truly fix, remove 'navbar-brand' from metrics div and use `navbar-metrics` 
    to properly style it */
 body[class^="vscode-"] .navbar > div > .navbar-text:not(.navbar-brand) > div > div > div:last-of-type {
-  margin-top: -10px;
   transform: scale(0.7);
 }
\ No newline at end of file