diff --git a/game/ato/flightstate/actionstate.py b/game/ato/flightstate/actionstate.py
index 15884c01..323ab709 100644
--- a/game/ato/flightstate/actionstate.py
+++ b/game/ato/flightstate/actionstate.py
@@ -21,5 +21,5 @@ class ActionState:
     def is_finished(self) -> bool:
         return self._finished
 
-    def on_game_tick(self, time: datetime, duration: timedelta) -> None:
-        self.action.update_state(self, time, duration)
+    def on_game_tick(self, time: datetime, duration: timedelta) -> timedelta:
+        return self.action.update_state(self, time, duration)
diff --git a/game/ato/flightstate/inflight.py b/game/ato/flightstate/inflight.py
index 3654255e..fc0be5e3 100644
--- a/game/ato/flightstate/inflight.py
+++ b/game/ato/flightstate/inflight.py
@@ -6,7 +6,6 @@ from datetime import datetime, timedelta
 from typing import TYPE_CHECKING
 
 from dcs import Point
-from dcs.task import Task
 
 from game.ato.flightstate import Completed
 from game.ato.flightstate.actionstate import ActionState
@@ -111,11 +110,12 @@ class InFlight(FlightState, ABC):
     def on_game_tick(
         self, events: GameUpdateEvents, time: datetime, duration: timedelta
     ) -> None:
-        if (action := self.current_action) is not None:
-            action.on_game_tick(time, duration)
+        while (action := self.current_action) is not None:
+            duration = action.on_game_tick(time, duration)
             if action.is_finished():
                 self.pending_actions.popleft()
-            return
+            if duration <= timedelta():
+                return
 
         self.elapsed_time += duration
         if self.elapsed_time > self.total_time_to_next_waypoint:
diff --git a/game/flightplan/waypointactions/hold.py b/game/flightplan/waypointactions/hold.py
index 9b2775e8..71da9227 100644
--- a/game/flightplan/waypointactions/hold.py
+++ b/game/flightplan/waypointactions/hold.py
@@ -32,9 +32,12 @@ class Hold(WaypointAction):
 
     def update_state(
         self, state: ActionState, time: datetime, duration: timedelta
-    ) -> None:
-        if self._push_time_provider() <= time:
+    ) -> timedelta:
+        push_time = self._push_time_provider()
+        if push_time <= time:
             state.finish()
+            return time - push_time
+        return timedelta()
 
     def iter_tasks(self, ctx: TaskContext) -> Iterator[Task]:
         remaining_time = self._push_time_provider() - ctx.mission_start_time
diff --git a/game/flightplan/waypointactions/waypointaction.py b/game/flightplan/waypointactions/waypointaction.py
index 52095bf5..a82a8678 100644
--- a/game/flightplan/waypointactions/waypointaction.py
+++ b/game/flightplan/waypointactions/waypointaction.py
@@ -21,7 +21,7 @@ class WaypointAction(ABC):
     @abstractmethod
     def update_state(
         self, state: ActionState, time: datetime, duration: timedelta
-    ) -> None:
+    ) -> timedelta:
         ...
 
     @abstractmethod
diff --git a/tests/ato/flightstate/test_actionstate.py b/tests/ato/flightstate/test_actionstate.py
index aeb4348b..b62864e7 100644
--- a/tests/ato/flightstate/test_actionstate.py
+++ b/tests/ato/flightstate/test_actionstate.py
@@ -14,8 +14,8 @@ class TestAction(WaypointAction):
 
     def update_state(
         self, state: ActionState, time: datetime, duration: timedelta
-    ) -> None:
-        pass
+    ) -> timedelta:
+        return timedelta()
 
     def iter_tasks(self, ctx: TaskContext) -> Iterator[Task]:
         yield from []
diff --git a/tests/flightplan/waypointactions/test_hold.py b/tests/flightplan/waypointactions/test_hold.py
index 3831569b..e5f86a1b 100644
--- a/tests/flightplan/waypointactions/test_hold.py
+++ b/tests/flightplan/waypointactions/test_hold.py
@@ -39,19 +39,21 @@ def test_hold_tick() -> None:
     t0 = datetime(1999, 3, 28)
     task = Hold(lambda: t0 + timedelta(minutes=5), meters(8000), kph(400))
     state = ActionState(task)
-    task.update_state(state, t0, timedelta())
+    assert not task.update_state(state, t0, timedelta())
     assert not state.is_finished()
-    task.update_state(state, t0 + timedelta(minutes=1), timedelta(minutes=1))
+    assert not task.update_state(state, t0 + timedelta(minutes=1), timedelta(minutes=1))
     assert not state.is_finished()
-    task.update_state(state, t0 + timedelta(minutes=2), timedelta(minutes=1))
+    assert not task.update_state(state, t0 + timedelta(minutes=2), timedelta(minutes=1))
     assert not state.is_finished()
-    task.update_state(state, t0 + timedelta(minutes=3), timedelta(minutes=1))
+    assert not task.update_state(state, t0 + timedelta(minutes=3), timedelta(minutes=1))
     assert not state.is_finished()
-    task.update_state(state, t0 + timedelta(minutes=4), timedelta(minutes=1))
+    assert not task.update_state(state, t0 + timedelta(minutes=4), timedelta(minutes=1))
     assert not state.is_finished()
-    task.update_state(state, t0 + timedelta(minutes=5), timedelta(minutes=1))
+    assert not task.update_state(state, t0 + timedelta(minutes=5), timedelta(minutes=1))
     assert state.is_finished()
-    task.update_state(state, t0 + timedelta(minutes=6), timedelta(minutes=1))
+    assert task.update_state(
+        state, t0 + timedelta(minutes=6), timedelta(minutes=1)
+    ) == timedelta(minutes=1)
     assert state.is_finished()