diff --git a/apis/config/v1beta1/configuration_types.go b/apis/config/v1beta1/configuration_types.go
index 2b75e2d1d4..dffc6bdd0c 100644
--- a/apis/config/v1beta1/configuration_types.go
+++ b/apis/config/v1beta1/configuration_types.go
@@ -211,6 +211,16 @@ type WaitForPodsReady struct {
 	// +optional
 	Timeout *metav1.Duration `json:"timeout,omitempty"`
 
+	// RecoveryTimeout defines an optional timeout, measured since the
+	// last transition to the PodsReady=false condition after a Workload is Admitted and running.
+	// Such a transition may happen when a Pod failed and the replacement Pod
+	// is awaited to be scheduled.
+	// After exceeding the timeout the corresponding job gets suspended again
+	// and requeued after the backoff delay. The timeout is enforced only if waitForPodsReady.enable=true.
+	// Defaults to 3 mins.
+	// +optional
+	RecoveryTimeout *metav1.Duration `json:"recoveryTimeout,omitempty"`
+
 	// BlockAdmission when true, cluster queue will block admissions for all
 	// subsequent jobs until the jobs reach the PodsReady=true condition.
 	// This setting is only honored when `Enable` is set to true.
diff --git a/apis/config/v1beta1/defaults.go b/apis/config/v1beta1/defaults.go
index 26fb9ce360..97e43ca063 100644
--- a/apis/config/v1beta1/defaults.go
+++ b/apis/config/v1beta1/defaults.go
@@ -41,6 +41,7 @@ const (
 	DefaultClientConnectionQPS                  float32 = 20.0
 	DefaultClientConnectionBurst                int32   = 30
 	defaultPodsReadyTimeout                             = 5 * time.Minute
+	defaultPodsRecoveryTimeout                          = 3 * time.Minute
 	DefaultQueueVisibilityUpdateIntervalSeconds int32   = 5
 	DefaultClusterQueuesMaxCount                int32   = 10
 	defaultJobFrameworkName                             = "batch/job"
@@ -119,6 +120,9 @@ func SetDefaults_Configuration(cfg *Configuration) {
 		if cfg.WaitForPodsReady.Timeout == nil {
 			cfg.WaitForPodsReady.Timeout = &metav1.Duration{Duration: defaultPodsReadyTimeout}
 		}
+		if cfg.WaitForPodsReady.RecoveryTimeout == nil {
+			cfg.WaitForPodsReady.RecoveryTimeout = &metav1.Duration{Duration: defaultPodsRecoveryTimeout}
+		}
 		if cfg.WaitForPodsReady.BlockAdmission == nil {
 			defaultBlockAdmission := true
 			if !cfg.WaitForPodsReady.Enable {
diff --git a/apis/config/v1beta1/defaults_test.go b/apis/config/v1beta1/defaults_test.go
index a0da4076ba..2de09a17eb 100644
--- a/apis/config/v1beta1/defaults_test.go
+++ b/apis/config/v1beta1/defaults_test.go
@@ -122,7 +122,8 @@ func TestSetDefaults_Configuration(t *testing.T) {
 		WorkerLostTimeout: &metav1.Duration{Duration: DefaultMultiKueueWorkerLostTimeout},
 	}
 
-	podsReadyTimeoutTimeout := metav1.Duration{Duration: defaultPodsReadyTimeout}
+	podsReadyTimeout := metav1.Duration{Duration: defaultPodsReadyTimeout}
+	podsReadyRecoveryTimeout := metav1.Duration{Duration: defaultPodsRecoveryTimeout}
 	podsReadyTimeoutOverwrite := metav1.Duration{Duration: time.Minute}
 
 	testCases := map[string]struct {
@@ -388,9 +389,10 @@ func TestSetDefaults_Configuration(t *testing.T) {
 			},
 			want: &Configuration{
 				WaitForPodsReady: &WaitForPodsReady{
-					Enable:         true,
-					BlockAdmission: ptr.To(true),
-					Timeout:        &podsReadyTimeoutTimeout,
+					Enable:          true,
+					BlockAdmission:  ptr.To(true),
+					Timeout:         &podsReadyTimeout,
+					RecoveryTimeout: &podsReadyRecoveryTimeout,
 					RequeuingStrategy: &RequeuingStrategy{
 						Timestamp:          ptr.To(EvictionTimestamp),
 						BackoffBaseSeconds: ptr.To[int32](DefaultRequeuingBackoffBaseSeconds),
@@ -420,9 +422,10 @@ func TestSetDefaults_Configuration(t *testing.T) {
 			},
 			want: &Configuration{
 				WaitForPodsReady: &WaitForPodsReady{
-					Enable:         false,
-					BlockAdmission: ptr.To(false),
-					Timeout:        &podsReadyTimeoutTimeout,
+					Enable:          false,
+					BlockAdmission:  ptr.To(false),
+					Timeout:         &podsReadyTimeout,
+					RecoveryTimeout: &podsReadyRecoveryTimeout,
 					RequeuingStrategy: &RequeuingStrategy{
 						Timestamp:          ptr.To(EvictionTimestamp),
 						BackoffBaseSeconds: ptr.To[int32](DefaultRequeuingBackoffBaseSeconds),
@@ -458,9 +461,10 @@ func TestSetDefaults_Configuration(t *testing.T) {
 			},
 			want: &Configuration{
 				WaitForPodsReady: &WaitForPodsReady{
-					Enable:         true,
-					BlockAdmission: ptr.To(true),
-					Timeout:        &podsReadyTimeoutOverwrite,
+					Enable:          true,
+					BlockAdmission:  ptr.To(true),
+					Timeout:         &podsReadyTimeoutOverwrite,
+					RecoveryTimeout: &podsReadyRecoveryTimeout,
 					RequeuingStrategy: &RequeuingStrategy{
 						Timestamp:          ptr.To(CreationTimestamp),
 						BackoffBaseSeconds: ptr.To[int32](63),
diff --git a/apis/config/v1beta1/zz_generated.deepcopy.go b/apis/config/v1beta1/zz_generated.deepcopy.go
index 3cd2aa1e5d..1c788bea7e 100644
--- a/apis/config/v1beta1/zz_generated.deepcopy.go
+++ b/apis/config/v1beta1/zz_generated.deepcopy.go
@@ -512,6 +512,11 @@ func (in *WaitForPodsReady) DeepCopyInto(out *WaitForPodsReady) {
 		*out = new(v1.Duration)
 		**out = **in
 	}
+	if in.RecoveryTimeout != nil {
+		in, out := &in.RecoveryTimeout, &out.RecoveryTimeout
+		*out = new(v1.Duration)
+		**out = **in
+	}
 	if in.BlockAdmission != nil {
 		in, out := &in.BlockAdmission, &out.BlockAdmission
 		*out = new(bool)
diff --git a/pkg/config/config_test.go b/pkg/config/config_test.go
index 0905837006..e40a17c540 100644
--- a/pkg/config/config_test.go
+++ b/pkg/config/config_test.go
@@ -629,6 +629,7 @@ webhook:
 					Enable:         true,
 					BlockAdmission: ptr.To(false),
 					Timeout:        &metav1.Duration{Duration: 50 * time.Second},
+					RecoveryTimeout: &metav1.Duration{Duration: 3 * time.Minute},
 					RequeuingStrategy: &configapi.RequeuingStrategy{
 						Timestamp:          ptr.To(configapi.CreationTimestamp),
 						BackoffLimitCount:  ptr.To[int32](10),
diff --git a/pkg/config/validation.go b/pkg/config/validation.go
index b65f282eb8..84e74da1a1 100644
--- a/pkg/config/validation.go
+++ b/pkg/config/validation.go
@@ -123,6 +123,10 @@ func validateWaitForPodsReady(c *configapi.Configuration) field.ErrorList {
 		allErrs = append(allErrs, field.Invalid(waitForPodsReadyPath.Child("timeout"),
 			c.WaitForPodsReady.Timeout, apimachineryvalidation.IsNegativeErrorMsg))
 	}
+	if c.WaitForPodsReady.RecoveryTimeout != nil && c.WaitForPodsReady.RecoveryTimeout.Duration < 0 {
+		allErrs = append(allErrs, field.Invalid(waitForPodsReadyPath.Child("recoveryTimeout"),
+			c.WaitForPodsReady.RecoveryTimeout, apimachineryvalidation.IsNegativeErrorMsg))
+	}
 	if strategy := c.WaitForPodsReady.RequeuingStrategy; strategy != nil {
 		if strategy.Timestamp != nil &&
 			*strategy.Timestamp != configapi.CreationTimestamp && *strategy.Timestamp != configapi.EvictionTimestamp {
diff --git a/pkg/config/validation_test.go b/pkg/config/validation_test.go
index ac503bda20..e6adfdf187 100644
--- a/pkg/config/validation_test.go
+++ b/pkg/config/validation_test.go
@@ -429,6 +429,23 @@ func TestValidate(t *testing.T) {
 				},
 			},
 		},
+		"negative waitForPodsReady.recoveryTimeout": {
+			cfg: &configapi.Configuration{
+				Integrations: defaultIntegrations,
+				WaitForPodsReady: &configapi.WaitForPodsReady{
+					Enable: true,
+					RecoveryTimeout: &metav1.Duration{
+						Duration: -1,
+					},
+				},
+			},
+			wantErr: field.ErrorList{
+				&field.Error{
+					Type:  field.ErrorTypeInvalid,
+					Field: "waitForPodsReady.recoveryTimeout",
+				},
+			},
+		},
 		"valid waitForPodsReady": {
 			cfg: &configapi.Configuration{
 				Integrations: defaultIntegrations,
@@ -437,6 +454,9 @@ func TestValidate(t *testing.T) {
 					Timeout: &metav1.Duration{
 						Duration: 50,
 					},
+					RecoveryTimeout: &metav1.Duration{
+						Duration: 5,
+					},
 					BlockAdmission: ptr.To(false),
 					RequeuingStrategy: &configapi.RequeuingStrategy{
 						Timestamp:          ptr.To(configapi.CreationTimestamp),
diff --git a/pkg/controller/core/core.go b/pkg/controller/core/core.go
index 3c6862f87b..b18ba91a8a 100644
--- a/pkg/controller/core/core.go
+++ b/pkg/controller/core/core.go
@@ -92,7 +92,8 @@ func waitForPodsReady(cfg *configapi.WaitForPodsReady) *waitForPodsReadyConfig {
 		return nil
 	}
 	result := waitForPodsReadyConfig{
-		timeout: cfg.Timeout.Duration,
+		timeout:         cfg.Timeout.Duration,
+		recoveryTimeout: cfg.RecoveryTimeout.Duration,
 	}
 	if cfg.RequeuingStrategy != nil {
 		result.requeuingBackoffBaseSeconds = *cfg.RequeuingStrategy.BackoffBaseSeconds
diff --git a/pkg/controller/core/workload_controller.go b/pkg/controller/core/workload_controller.go
index 8d502d2927..1765e13fa0 100644
--- a/pkg/controller/core/workload_controller.go
+++ b/pkg/controller/core/workload_controller.go
@@ -62,6 +62,7 @@ var (
 
 type waitForPodsReadyConfig struct {
 	timeout                     time.Duration
+	recoveryTimeout             time.Duration
 	requeuingBackoffLimitCount  *int32
 	requeuingBackoffBaseSeconds int32
 	requeuingBackoffMaxDuration time.Duration
diff --git a/site/content/en/docs/reference/kueue-config.v1beta1.md b/site/content/en/docs/reference/kueue-config.v1beta1.md
index a2e17e6043..49fdcd5dd5 100644
--- a/site/content/en/docs/reference/kueue-config.v1beta1.md
+++ b/site/content/en/docs/reference/kueue-config.v1beta1.md
@@ -902,6 +902,19 @@ evicted and requeued in the same cluster queue.
 Defaults to 5min.</p>
 </td>
 </tr>
+<tr><td><code>recoveryTimeout</code><br/>
+<a href="https://kubernetes.io/docs/reference/generated/kubernetes-api/v1.28/#duration-v1-meta"><code>k8s.io/apimachinery/pkg/apis/meta/v1.Duration</code></a>
+</td>
+<td>
+   <p>RecoveryTimeout defines an optional timeout, measured since the
+last transition to the PodsReady=false condition after a Workload is Admitted and running.
+Such a transition may happen when a Pod failed and the replacement Pod
+is awaited to be scheduled.
+After exceeding the timeout the corresponding job gets suspended again
+and requeued after the backoff delay. The timeout is enforced only if waitForPodsReady.enable=true.
+Defaults to 3 mins.</p>
+</td>
+</tr>
 <tr><td><code>blockAdmission</code> <B>[Required]</B><br/>
 <code>bool</code>
 </td>