diff --git a/staging/src/k8s.io/apiserver/pkg/server/filters/priority-and-fairness.go b/staging/src/k8s.io/apiserver/pkg/server/filters/priority-and-fairness.go
index decd9d6ca82..5986993d4bb 100644
--- a/staging/src/k8s.io/apiserver/pkg/server/filters/priority-and-fairness.go
+++ b/staging/src/k8s.io/apiserver/pkg/server/filters/priority-and-fairness.go
@@ -52,8 +52,8 @@ var waitingMark = &requestWatermark{
 	phase: epmetrics.WaitingPhase,
 }
 
-var atomicMutatingExecuting, atomicReadOnlyExecuting int32
-var atomicMutatingWaiting, atomicReadOnlyWaiting int32
+var atomicMutatingExecuting, atomicReadOnlyExecuting atomic.Int32
+var atomicMutatingWaiting, atomicReadOnlyWaiting atomic.Int32
 
 // newInitializationSignal is defined for testing purposes.
 var newInitializationSignal = utilflowcontrol.NewInitializationSignal
@@ -143,16 +143,16 @@ func (h *priorityAndFairnessHandler) Handle(w http.ResponseWriter, r *http.Reque
 	isMutatingRequest := !nonMutatingRequestVerbs.Has(requestInfo.Verb)
 	noteExecutingDelta := func(delta int32) {
 		if isMutatingRequest {
-			watermark.recordMutating(int(atomic.AddInt32(&atomicMutatingExecuting, delta)))
+			watermark.recordMutating(int(atomicMutatingExecuting.Add(delta)))
 		} else {
-			watermark.recordReadOnly(int(atomic.AddInt32(&atomicReadOnlyExecuting, delta)))
+			watermark.recordReadOnly(int(atomicReadOnlyExecuting.Add(delta)))
 		}
 	}
 	noteWaitingDelta := func(delta int32) {
 		if isMutatingRequest {
-			waitingMark.recordMutating(int(atomic.AddInt32(&atomicMutatingWaiting, delta)))
+			waitingMark.recordMutating(int(atomicMutatingWaiting.Add(delta)))
 		} else {
-			waitingMark.recordReadOnly(int(atomic.AddInt32(&atomicReadOnlyWaiting, delta)))
+			waitingMark.recordReadOnly(int(atomicReadOnlyWaiting.Add(delta)))
 		}
 	}
 	queueNote := func(inQueue bool) {
diff --git a/staging/src/k8s.io/apiserver/pkg/server/filters/priority-and-fairness_test.go b/staging/src/k8s.io/apiserver/pkg/server/filters/priority-and-fairness_test.go
index 55ef87fdf36..18d99292ccd 100644
--- a/staging/src/k8s.io/apiserver/pkg/server/filters/priority-and-fairness_test.go
+++ b/staging/src/k8s.io/apiserver/pkg/server/filters/priority-and-fairness_test.go
@@ -128,20 +128,25 @@ func newApfServerWithSingleRequest(t *testing.T, decision mockDecision) *httptes
 			t.Errorf("execute should not be invoked")
 		}
 		// atomicReadOnlyExecuting can be either 0 or 1 as we test one request at a time.
-		if decision != decisionSkipFilter && atomicReadOnlyExecuting != 1 {
-			t.Errorf("Wanted %d requests executing, got %d", 1, atomicReadOnlyExecuting)
+		currentValue := atomicReadOnlyExecuting.Load()
+		if decision != decisionSkipFilter && currentValue != 1 {
+			t.Errorf("Wanted %d requests executing, got %d", 1, currentValue)
 		}
 	}
+
 	postExecuteFunc := func() {}
 	// atomicReadOnlyWaiting can be either 0 or 1 as we test one request at a time.
 	postEnqueueFunc := func() {
-		if atomicReadOnlyWaiting != 1 {
-			t.Errorf("Wanted %d requests in queue, got %d", 1, atomicReadOnlyWaiting)
+		currentValue := atomicReadOnlyWaiting.Load()
+		if currentValue != 1 {
+			t.Errorf("Wanted %d requests in queue, got %d", 1, currentValue)
 		}
 	}
+
 	postDequeueFunc := func() {
-		if atomicReadOnlyWaiting != 0 {
-			t.Errorf("Wanted %d requests in queue, got %d", 0, atomicReadOnlyWaiting)
+		currentValue := atomicReadOnlyWaiting.Load()
+		if currentValue != 0 {
+			t.Errorf("Wanted %d requests in queue, got %d", 0, currentValue)
 		}
 	}
 	return newApfServerWithHooks(t, decision, onExecuteFunc, postExecuteFunc, postEnqueueFunc, postDequeueFunc)
@@ -185,8 +190,9 @@ func newApfHandlerWithFilter(t *testing.T, flowControlFilter utilflowcontrol.Int
 			// TODO: all test(s) using this filter must run
 			// serially to each other
 			defer func() {
-				if atomicReadOnlyExecuting != 0 {
-					t.Errorf("Wanted %d requests executing, got %d", 0, atomicReadOnlyExecuting)
+				currentValue := atomicReadOnlyExecuting.Load()
+				if currentValue != 0 {
+					t.Errorf("Wanted %d requests executing, got %d", 0, currentValue)
 				}
 			}()
 			apfHandler.ServeHTTP(w, r)
@@ -280,8 +286,9 @@ func TestApfExecuteMultipleRequests(t *testing.T) {
 	onExecuteFunc := func() {
 		preStartExecute.Done()
 		preStartExecute.Wait()
-		if int(atomicReadOnlyExecuting) != concurrentRequests {
-			t.Errorf("Wanted %d requests executing, got %d", concurrentRequests, atomicReadOnlyExecuting)
+		currentValue := atomicReadOnlyExecuting.Load()
+		if int(currentValue) != concurrentRequests {
+			t.Errorf("Wanted %d requests executing, got %d", concurrentRequests, currentValue)
 		}
 		postStartExecute.Done()
 		postStartExecute.Wait()
@@ -290,9 +297,9 @@ func TestApfExecuteMultipleRequests(t *testing.T) {
 	postEnqueueFunc := func() {
 		preEnqueue.Done()
 		preEnqueue.Wait()
-		if int(atomicReadOnlyWaiting) != concurrentRequests {
-			t.Errorf("Wanted %d requests in queue, got %d", 1, atomicReadOnlyWaiting)
-
+		currentValue := atomicReadOnlyWaiting.Load()
+		if int(currentValue) != concurrentRequests {
+			t.Errorf("Wanted %d requests in queue, got %d", concurrentRequests, currentValue)
 		}
 		postEnqueue.Done()
 		postEnqueue.Wait()
@@ -301,8 +308,9 @@ func TestApfExecuteMultipleRequests(t *testing.T) {
 	postDequeueFunc := func() {
 		preDequeue.Done()
 		preDequeue.Wait()
-		if atomicReadOnlyWaiting != 0 {
-			t.Errorf("Wanted %d requests in queue, got %d", 0, atomicReadOnlyWaiting)
+		currentValue := atomicReadOnlyWaiting.Load()
+		if currentValue != 0 {
+			t.Errorf("Wanted %d requests in queue, got %d", 0, currentValue)
 		}
 		postDequeue.Done()
 		postDequeue.Wait()