diff --git a/dev-packages/e2e-tests/maestro/captureReplay.yml b/dev-packages/e2e-tests/maestro/captureReplay.yml
index 9499f427c2..6833d82f0b 100644
--- a/dev-packages/e2e-tests/maestro/captureReplay.yml
+++ b/dev-packages/e2e-tests/maestro/captureReplay.yml
@@ -6,15 +6,11 @@ jsEngine: graaljs
     env:
       replaysOnErrorSampleRate: 1.0
 # Prime the native replay buffer before capturing the exception so a replay_id
-# is attached to the event (iOS only; Android does not reliably capture replays
-# in CI — see https://github.com/getsentry/sentry-react-native/pull/4277).
-- runFlow:
-    file: utils/primeReplayBuffer.yml
-    when:
-      platform: iOS
+# is attached to the event. Runs on both platforms: Android historically did not
+# reliably capture replays in CI (see
+# https://github.com/getsentry/sentry-react-native/pull/4277); this re-evaluates
+# that with deterministic buffer priming.
+- runFlow: utils/primeReplayBuffer.yml
 - tapOn: "Capture Exception"
 - runFlow: utils/assertEventIdVisible.yml
-- runFlow:
-    file: utils/assertReplay.yml
-    when:
-      platform: iOS
+- runFlow: utils/assertReplay.yml
diff --git a/dev-packages/e2e-tests/maestro/utils/assertReplay.yml b/dev-packages/e2e-tests/maestro/utils/assertReplay.yml
index b9885b8227..e21f10f3fe 100644
--- a/dev-packages/e2e-tests/maestro/utils/assertReplay.yml
+++ b/dev-packages/e2e-tests/maestro/utils/assertReplay.yml
@@ -20,4 +20,7 @@ jsEngine: graaljs
 - assertTrue: ${output.replayId}
 - assertTrue: ${output.replayDuration}
 - assertTrue: ${output.replaySegments}
-- assertTrue: ${output.replayCodec == "ftypmp42"}
+# Assert a valid MP4 container was produced (the "ftyp" box at byte offset 4)
+# rather than a platform-specific major brand: iOS (AVAssetWriter) emits "mp42"
+# while Android (MediaMuxer) may emit a different brand such as "isom".
+- assertTrue: ${output.replayCodec.startsWith("ftyp")}