diff --git a/src/prefect/engine.py b/src/prefect/engine.py
index ef3fa7e790ea..ec9f286cb3d0 100644
--- a/src/prefect/engine.py
+++ b/src/prefect/engine.py
@@ -167,6 +167,7 @@
 from prefect.settings import (
     PREFECT_DEBUG_MODE,
     PREFECT_EXPERIMENTAL_ENABLE_NEW_ENGINE,
+    PREFECT_RUN_ON_COMPLETION_HOOKS_ON_CACHED,
     PREFECT_TASK_INTROSPECTION_WARN_THRESHOLD,
     PREFECT_TASKS_REFRESH_CACHE,
     PREFECT_UI_URL,
@@ -2117,6 +2118,19 @@ async def tick():
     # flag to ensure we only update the task run name once
     run_name_set = False
 
+    run_on_completion_hooks_on_cached = (
+        PREFECT_RUN_ON_COMPLETION_HOOKS_ON_CACHED
+        and state.is_completed()
+        and state.name == "Cached"
+    )
+
+    if run_on_completion_hooks_on_cached:
+        await _run_task_hooks(
+            task=task,
+            task_run=task_run,
+            state=state,
+        )
+
     # Only run the task if we enter a `RUNNING` state
     while state.is_running():
         # Retrieve the latest metadata for the task run context
@@ -2326,9 +2340,16 @@ async def _run_task_hooks(task: Task, task_run: TaskRun, state: State) -> None:
     catch and log any errors that occur.
     """
     hooks = None
+    run_on_completion_hooks_on_cached = (
+        PREFECT_RUN_ON_COMPLETION_HOOKS_ON_CACHED
+        and state.is_completed()
+        and state.name == "Cached"
+    )
     if state.is_failed() and task.on_failure:
         hooks = task.on_failure
-    elif state.is_completed() and task.on_completion:
+    elif (
+        state.is_completed() or run_on_completion_hooks_on_cached
+    ) and task.on_completion:
         hooks = task.on_completion
 
     if hooks:
diff --git a/src/prefect/settings.py b/src/prefect/settings.py
index ec2b42491624..7130ca65e20a 100644
--- a/src/prefect/settings.py
+++ b/src/prefect/settings.py
@@ -1759,6 +1759,11 @@ def default_cloud_ui_url(settings, value):
 How long to cache related resource data for emitting server-side vents
 """
 
+PREFECT_RUN_ON_COMPLETION_HOOKS_ON_CACHED = Setting(bool, default=False)
+"""
+Whether or not to run on_completion hooks on cached task runs.
+"""
+
 
 def automation_settings_enabled() -> bool:
     """
diff --git a/tests/test_tasks.py b/tests/test_tasks.py
index 5d3b6e92ba1a..849660b95714 100644
--- a/tests/test_tasks.py
+++ b/tests/test_tasks.py
@@ -29,6 +29,7 @@
 from prefect.settings import (
     PREFECT_DEBUG_MODE,
     PREFECT_EXPERIMENTAL_ENABLE_NEW_ENGINE,
+    PREFECT_RUN_ON_COMPLETION_HOOKS_ON_CACHED,
     PREFECT_TASK_DEFAULT_RETRIES,
     PREFECT_TASKS_REFRESH_CACHE,
     temporary_settings,
@@ -3767,6 +3768,90 @@ def my_flow():
         assert my_mock.call_args_list == [call(), call()]
 
 
+class TestOnCompletionHooksForCachedTasks:
+    @pytest.mark.parametrize("run_hooks_on_cached", [True, False])
+    def test_on_completion_hooks_for_cached_tasks(self, run_hooks_on_cached):
+        my_mock = MagicMock()
+
+        def completed1(task, task_run, state):
+            my_mock("completed1")
+
+        def completed2(task, task_run, state):
+            my_mock("completed2")
+
+        @task(
+            cache_key_fn=lambda *_: "cache_2",
+            cache_expiration=datetime.timedelta(seconds=5),
+            on_completion=[completed1, completed2],
+        )
+        def cached_task():
+            return "cached_result"
+
+        @flow
+        def test_flow():
+            return cached_task._run()
+
+        with temporary_settings(
+            {PREFECT_RUN_ON_COMPLETION_HOOKS_ON_CACHED: run_hooks_on_cached}
+        ):
+            # First run to cache the result
+            state = test_flow()
+            assert my_mock.call_args_list == [call("completed1"), call("completed2")]
+            assert not (state.is_completed() and state.name == "Cached")
+
+            my_mock.reset_mock()
+
+            # Second run to test the hooks behavior
+            state = test_flow()
+            assert state.is_completed() and state.name == "Cached"
+            if run_hooks_on_cached:
+                assert my_mock.call_args_list == [
+                    call("completed1"),
+                    call("completed2"),
+                ]
+            else:
+                assert my_mock.call_args_list == []
+
+    @pytest.mark.parametrize("run_hooks_on_cached", [True, False])
+    def test_exception_in_on_completion_hook_for_cached_task(
+        self, caplog, run_hooks_on_cached
+    ):
+        def failing_hook(task, task_run, state):
+            raise ValueError("Hook failed")
+
+        @task(
+            cache_key_fn=lambda *_: "cache_3",
+            cache_expiration=datetime.timedelta(seconds=5),
+            on_completion=[failing_hook],
+        )
+        def cached_task():
+            return "cached_result"
+
+        @flow
+        def test_flow():
+            return cached_task._run()
+
+        with temporary_settings(
+            {PREFECT_RUN_ON_COMPLETION_HOOKS_ON_CACHED: run_hooks_on_cached}
+        ):
+            # First run to cache the result
+            state = test_flow()
+            assert state.result() == "cached_result"
+            assert not (state.is_completed() and state.name == "Cached")
+
+            caplog.clear()
+
+            # Second run to test the hook behavior
+            state = test_flow()
+            assert state.is_completed() and state.name == "Cached"
+            assert state.result() == "cached_result"
+
+            if run_hooks_on_cached:
+                assert "ValueError: Hook failed" in caplog.text
+            else:
+                assert "ValueError: Hook failed" not in caplog.text
+
+
 class TestTaskHooksOnFailure:
     def test_noniterable_hook_raises(self):
         def failure_hook():