--- a/tools/profiler/core/platform.cpp
+++ b/tools/profiler/core/platform.cpp
@@ -2540,18 +2540,16 @@ static void locked_profiler_stream_json_
       profiledThreadData->StreamJSON(
           buffer, cx, aWriter, CorePS::ProcessName(aLock),
           CorePS::ProcessStartTime(), aSinceTime,
           ActivePS::FeatureJSTracer(aLock), aService);
     }
 
 #if defined(GP_OS_android)
     if (ActivePS::FeatureJava(aLock)) {
-      java::GeckoJavaSampler::Pause();
-
       // We are allocating it chunk by chunk. So this will not allocate 64 MiB
       // at once. This size should be more than enough for java threads.
       // This buffer is being created for each process but Android has
       // relatively less processes compared to desktop, so it's okay here.
       mozilla::ProfileBufferChunkManagerWithLocalLimit chunkManager(
           64 * 1024 * 1024, 1024 * 1024);
       ProfileChunkedBuffer bufferManager(
           ProfileChunkedBuffer::ThreadSafety::WithoutMutex, chunkManager);
@@ -2562,18 +2560,16 @@ static void locked_profiler_stream_json_
       // java thread, we have to get thread id and name via JNI.
       RefPtr<ThreadInfo> threadInfo = new ThreadInfo(
           "Java Main Thread", 0, false, CorePS::ProcessStartTime());
       ProfiledThreadData profiledThreadData(threadInfo, nullptr);
       profiledThreadData.StreamJSON(javaBuffer, nullptr, aWriter,
                                     CorePS::ProcessName(aLock),
                                     CorePS::ProcessStartTime(), aSinceTime,
                                     ActivePS::FeatureJSTracer(aLock), nullptr);
-
-      java::GeckoJavaSampler::Unpause();
     }
 #endif
 
     UniquePtr<char[]> baseProfileThreads =
         ActivePS::MoveBaseProfileThreads(aLock);
     if (baseProfileThreads) {
       aWriter.Splice(baseProfileThreads.get());
     }