diff --git a/pkg/controllers/node/termination/controller.go b/pkg/controllers/node/termination/controller.go
index b8e49a3931..f51bcafa05 100644
--- a/pkg/controllers/node/termination/controller.go
+++ b/pkg/controllers/node/termination/controller.go
@@ -28,6 +28,7 @@ import (
 	storagev1 "k8s.io/api/storage/v1"
 	"k8s.io/apimachinery/pkg/api/equality"
 	"k8s.io/apimachinery/pkg/api/errors"
+	"k8s.io/apimachinery/pkg/types"
 	"k8s.io/apimachinery/pkg/util/sets"
 	"k8s.io/client-go/util/workqueue"
 	"k8s.io/utils/clock"
@@ -193,6 +194,15 @@ func (c *Controller) ensureVolumesDetached(ctx context.Context, node *corev1.Nod
 	return len(filteredVolumeAttachments) == 0, nil
 }
 
+// volumeIsNotAttachedOrBound returns true if the persistent volume storage is neither Attached nor Bound
+func volumeIsNotAttachedOrBound(ctx context.Context, kubeClient client.Client, pvName string) (bool, error) {
+	pv := &corev1.PersistentVolume{}
+	if err := kubeClient.Get(ctx, types.NamespacedName{Name: pvName}, pv); err != nil {
+		return false, fmt.Errorf("getting persistent volume %q, %w", pvName, err)
+	}
+	return pv.Status.Phase == corev1.VolumePending || pv.Status.Phase == corev1.VolumeReleased || pv.Status.Phase == corev1.VolumeFailed, nil
+}
+
 // filterVolumeAttachments filters out storagev1.VolumeAttachments that should not block the termination
 // of the passed corev1.Node
 func filterVolumeAttachments(ctx context.Context, kubeClient client.Client, node *corev1.Node, volumeAttachments []*storagev1.VolumeAttachment, clk clock.Clock) ([]*storagev1.VolumeAttachment, error) {
@@ -225,6 +235,21 @@ func filterVolumeAttachments(ctx context.Context, kubeClient client.Client, node
 			}
 		}
 	}
+
+	// ignore volume attachments associated with volumes that are no longer bound
+	for _, va := range volumeAttachments {
+		unattached, err := volumeIsNotAttachedOrBound(ctx, kubeClient, *va.Spec.Source.PersistentVolumeName)
+		if errors.IsNotFound(err) {
+			continue
+		}
+		if err != nil {
+			return nil, err
+		}
+		if unattached {
+			shouldFilterOutVolume.Insert(*va.Spec.Source.PersistentVolumeName)
+		}
+	}
+
 	filteredVolumeAttachments := lo.Reject(volumeAttachments, func(v *storagev1.VolumeAttachment, _ int) bool {
 		pvName := v.Spec.Source.PersistentVolumeName
 		return pvName == nil || shouldFilterOutVolume.Has(*pvName)
diff --git a/pkg/controllers/node/termination/suite_test.go b/pkg/controllers/node/termination/suite_test.go
index 32151e311f..5be532e8f6 100644
--- a/pkg/controllers/node/termination/suite_test.go
+++ b/pkg/controllers/node/termination/suite_test.go
@@ -800,6 +800,25 @@ var _ = Describe("Termination", func() {
 				ExpectObjectReconciled(ctx, env.Client, terminationController, node)
 				ExpectNotFound(ctx, env.Client, node)
 			})
+			It("should not wait for volume attachments with released persistent volumes", func() {
+				va := test.VolumeAttachment(test.VolumeAttachmentOptions{
+					NodeName:   node.Name,
+					VolumeName: "foo",
+				})
+				pv := test.PersistentVolume(test.PersistentVolumeOptions{
+					ObjectMeta: metav1.ObjectMeta{
+						Name: "foo",
+					},
+				})
+				pv.Status.Phase = corev1.VolumeReleased
+
+				ExpectApplied(ctx, env.Client, node, nodeClaim, nodePool, pv, va)
+				Expect(env.Client.Delete(ctx, node)).To(Succeed())
+
+				ExpectObjectReconciled(ctx, env.Client, terminationController, node)
+				ExpectObjectReconciled(ctx, env.Client, terminationController, node)
+				ExpectNotFound(ctx, env.Client, node)
+			})
 			It("should wait for volume attachments until the nodeclaim's termination grace period expires", func() {
 				va := test.VolumeAttachment(test.VolumeAttachmentOptions{
 					NodeName:   node.Name,