diff --git a/test/e2e/common/node/container_probe.go b/test/e2e/common/node/container_probe.go
index efc0455ecce..373b0658e10 100644
--- a/test/e2e/common/node/container_probe.go
+++ b/test/e2e/common/node/container_probe.go
@@ -235,6 +235,10 @@ var _ = SIGDescribe("Probing container", func() {
 		Description: A Pod is created with readiness probe with a Exec action on the Pod. If the readiness probe call does not return within the timeout specified, readiness probe MUST not be Ready.
 	*/
 	ginkgo.It("should not be ready with an exec readiness probe timeout [MinimumKubeletVersion:1.20] [NodeConformance]", func() {
+		// The ExecProbeTimeout feature gate exists to allow backwards-compatibility with pre-1.20 cluster behaviors, where readiness probe timeouts were ignored
+		// If ExecProbeTimeout feature gate is disabled, timeout enforcement for exec readiness probe is disabled, so we should skip this test
+		e2eskipper.SkipUnlessExecProbeTimeoutEnabled()
+
 		cmd := []string{"/bin/sh", "-c", "sleep 600"}
 		readinessProbe := &v1.Probe{
 			Handler:             execHandler([]string{"/bin/sh", "-c", "sleep 10"}),
@@ -255,6 +259,8 @@ var _ = SIGDescribe("Probing container", func() {
 		// The ExecProbeTimeout feature gate exists to allow backwards-compatibility with pre-1.20 cluster behaviors using dockershim, where livenessProbe timeouts were ignored
 		// If ExecProbeTimeout feature gate is disabled on a dockershim cluster, timeout enforcement for exec livenessProbes is disabled, but a failing liveness probe MUST still trigger a restart
 		// Note ExecProbeTimeout=false is not recommended for non-dockershim clusters (e.g., containerd), and this test will fail if run against such a configuration
+		e2eskipper.SkipUnlessExecProbeTimeoutEnabled()
+
 		cmd := []string{"/bin/sh", "-c", "sleep 600"}
 		livenessProbe := &v1.Probe{
 			Handler:             execHandler([]string{"/bin/sh", "-c", "sleep 10 & exit 1"}),