diff --git a/Lib/test/test_threading.py b/Lib/test/test_threading.py
index b8d310e4c99ff8..731e1046c000af 100644
--- a/Lib/test/test_threading.py
+++ b/Lib/test/test_threading.py
@@ -1381,6 +1381,33 @@ def test_native_id_after_fork(self):
         self.assertEqual(len(native_ids), 2)
         self.assertNotEqual(native_ids[0], native_ids[1])
 
+    def test_stop_the_world_during_finalization(self):
+        # gh-137433: Test functions that trigger a stop-the-world in the free
+        # threading build concurrent with interpreter finalization.
+        script = """if True:
+            import gc
+            import sys
+            import threading
+            NUM_THREADS = 5
+            b = threading.Barrier(NUM_THREADS + 1)
+            def run_in_bg():
+                b.wait()
+                while True:
+                    sys.setprofile(None)
+                    gc.collect()
+
+            for _ in range(NUM_THREADS):
+                t = threading.Thread(target=run_in_bg, daemon=True)
+                t.start()
+
+            b.wait()
+            print("Exiting...")
+        """
+        rc, out, err = assert_python_ok('-c', script)
+        self.assertEqual(rc, 0)
+        self.assertEqual(err, b"")
+        self.assertEqual(out.strip(), b"Exiting...")
+
 class ThreadJoinOnShutdown(BaseTestCase):
 
     def _run_and_join(self, script):
diff --git a/Misc/NEWS.d/next/Core_and_Builtins/2025-08-13-13-39-02.gh-issue-137433.g6Atfz.rst b/Misc/NEWS.d/next/Core_and_Builtins/2025-08-13-13-39-02.gh-issue-137433.g6Atfz.rst
new file mode 100644
index 00000000000000..0389cb0c735128
--- /dev/null
+++ b/Misc/NEWS.d/next/Core_and_Builtins/2025-08-13-13-39-02.gh-issue-137433.g6Atfz.rst
@@ -0,0 +1,3 @@
+Fix a potential deadlock in the :term:`free threading` build when daemon
+threads enable or disable profiling or tracing while the main thread is
+shutting down the interpreter.
diff --git a/Python/parking_lot.c b/Python/parking_lot.c
index e896dea02712f2..0ffea97b079943 100644
--- a/Python/parking_lot.c
+++ b/Python/parking_lot.c
@@ -342,7 +342,19 @@ _PyParkingLot_Park(const void *addr, const void *expected, size_t size,
     enqueue(bucket, addr, &wait);
     _PyRawMutex_Unlock(&bucket->mutex);
 
-    int res = _PySemaphore_Wait(&wait.sema, timeout_ns, detach);
+    PyThreadState *tstate = NULL;
+    if (detach) {
+        tstate = _PyThreadState_GET();
+        if (tstate && _PyThreadState_IsAttached(tstate)) {
+            // Only detach if we are attached
+            PyEval_ReleaseThread(tstate);
+        }
+        else {
+            tstate = NULL;
+        }
+    }
+
+    int res = _PySemaphore_Wait(&wait.sema, timeout_ns, 0);
     if (res == Py_PARK_OK) {
         goto done;
     }
@@ -354,7 +366,7 @@ _PyParkingLot_Park(const void *addr, const void *expected, size_t size,
         // Another thread has started to unpark us. Wait until we process the
         // wakeup signal.
         do {
-            res = _PySemaphore_Wait(&wait.sema, -1, detach);
+            res = _PySemaphore_Wait(&wait.sema, -1, 0);
         } while (res != Py_PARK_OK);
         goto done;
     }
@@ -366,6 +378,9 @@ _PyParkingLot_Park(const void *addr, const void *expected, size_t size,
 
 done:
     _PySemaphore_Destroy(&wait.sema);
+    if (tstate) {
+        PyEval_AcquireThread(tstate);
+    }
     return res;
 
 }
diff --git a/Python/pystate.c b/Python/pystate.c
index b7da4e981ac1eb..6859dbc9a7f469 100644
--- a/Python/pystate.c
+++ b/Python/pystate.c
@@ -2345,13 +2345,15 @@ stop_the_world(struct _stoptheworld_state *stw)
 {
     _PyRuntimeState *runtime = &_PyRuntime;
 
-    PyMutex_Lock(&stw->mutex);
+    // gh-137433: Acquire the rwmutex first to avoid deadlocks with daemon
+    // threads that may hang when blocked on lock acquisition.
     if (stw->is_global) {
         _PyRWMutex_Lock(&runtime->stoptheworld_mutex);
     }
     else {
         _PyRWMutex_RLock(&runtime->stoptheworld_mutex);
     }
+    PyMutex_Lock(&stw->mutex);
 
     HEAD_LOCK(runtime);
     stw->requested = 1;
@@ -2417,13 +2419,13 @@ start_the_world(struct _stoptheworld_state *stw)
     }
     stw->requester = NULL;
     HEAD_UNLOCK(runtime);
+    PyMutex_Unlock(&stw->mutex);
     if (stw->is_global) {
         _PyRWMutex_Unlock(&runtime->stoptheworld_mutex);
     }
     else {
         _PyRWMutex_RUnlock(&runtime->stoptheworld_mutex);
     }
-    PyMutex_Unlock(&stw->mutex);
 }
 #endif  // Py_GIL_DISABLED