diff --git a/cluster-autoscaler/processors/provreq/injector.go b/cluster-autoscaler/processors/provreq/injector.go
index 023a2b202c70..4217951f49a7 100644
--- a/cluster-autoscaler/processors/provreq/injector.go
+++ b/cluster-autoscaler/processors/provreq/injector.go
@@ -71,17 +71,6 @@ func (p *ProvisioningRequestPodsInjector) IsAvailableForProvisioning(pr *provreq
 	return true
 }
 
-// MarkAsAccepted marks the ProvisioningRequest as accepted.
-func (p *ProvisioningRequestPodsInjector) MarkAsAccepted(pr *provreqwrapper.ProvisioningRequest) error {
-	provreqconditions.AddOrUpdateCondition(pr, v1.Accepted, metav1.ConditionTrue, provreqconditions.AcceptedReason, provreqconditions.AcceptedMsg, metav1.NewTime(p.clock.Now()))
-	if _, err := p.client.UpdateProvisioningRequest(pr.ProvisioningRequest); err != nil {
-		klog.Errorf("failed add Accepted condition to ProvReq %s/%s, err: %v", pr.Namespace, pr.Name, err)
-		return err
-	}
-	p.lastProvisioningRequestProcessTime = p.clock.Now()
-	return nil
-}
-
 // MarkAsFailed marks the ProvisioningRequest as failed.
 func (p *ProvisioningRequestPodsInjector) MarkAsFailed(pr *provreqwrapper.ProvisioningRequest, reason string, message string) {
 	provreqconditions.AddOrUpdateCondition(pr, v1.Failed, metav1.ConditionTrue, reason, message, metav1.NewTime(p.clock.Now()))
@@ -120,9 +109,6 @@ func (p *ProvisioningRequestPodsInjector) GetPodsFromNextRequest(
 			p.MarkAsFailed(pr, provreqconditions.FailedToCreatePodsReason, err.Error())
 			continue
 		}
-		if err := p.MarkAsAccepted(pr); err != nil {
-			continue
-		}
 
 		return podsFromProvReq, nil
 	}
diff --git a/cluster-autoscaler/processors/provreq/injector_test.go b/cluster-autoscaler/processors/provreq/injector_test.go
index 0e51026c254c..d9d45720cc2b 100644
--- a/cluster-autoscaler/processors/provreq/injector_test.go
+++ b/cluster-autoscaler/processors/provreq/injector_test.go
@@ -21,7 +21,6 @@ import (
 	"testing"
 	"time"
 
-	apimeta "k8s.io/apimachinery/pkg/api/meta"
 	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
 	v1 "k8s.io/autoscaler/cluster-autoscaler/apis/provisioningrequest/autoscaling.x-k8s.io/v1"
 	"k8s.io/autoscaler/cluster-autoscaler/provisioningrequest/provreqclient"
@@ -83,25 +82,21 @@ func TestProvisioningRequestPodsInjector(t *testing.T) {
 		provReqs                         []*provreqwrapper.ProvisioningRequest
 		existingUnsUnschedulablePodCount int
 		wantUnscheduledPodCount          int
-		wantUpdatedConditionName         string
 	}{
 		{
 			name:                     "New ProvisioningRequest, pods are injected and Accepted condition is added",
 			provReqs:                 []*provreqwrapper.ProvisioningRequest{newProvReqA, provisionedAcceptedProvReqB},
 			wantUnscheduledPodCount:  podsA,
-			wantUpdatedConditionName: newProvReqA.Name,
 		},
 		{
 			name:                     "New ProvisioningRequest, pods are injected and Accepted condition is updated",
 			provReqs:                 []*provreqwrapper.ProvisioningRequest{newAcceptedProvReqA, provisionedAcceptedProvReqB},
 			wantUnscheduledPodCount:  podsA,
-			wantUpdatedConditionName: newAcceptedProvReqA.Name,
 		},
 		{
 			name:                     "Provisioned=False, pods are injected",
 			provReqs:                 []*provreqwrapper.ProvisioningRequest{notProvisionedAcceptedProvReqB, failedProvReq},
 			wantUnscheduledPodCount:  podsB,
-			wantUpdatedConditionName: notProvisionedAcceptedProvReqB.Name,
 		},
 		{
 			name:     "Provisioned=True, no pods are injected",
@@ -124,7 +119,6 @@ func TestProvisioningRequestPodsInjector(t *testing.T) {
 			provReqs:                         []*provreqwrapper.ProvisioningRequest{newProvReqA},
 			existingUnsUnschedulablePodCount: 50,
 			wantUnscheduledPodCount:          podsA + 50,
-			wantUpdatedConditionName:         newProvReqA.Name,
 		},
 	}
 	for _, tc := range testCases {
@@ -139,14 +133,6 @@ func TestProvisioningRequestPodsInjector(t *testing.T) {
 		if len(getUnscheduledPods) != tc.wantUnscheduledPodCount {
 			t.Errorf("%s failed: injector.Process return %d unscheduled pods, want %d", tc.name, len(getUnscheduledPods), tc.wantUnscheduledPodCount)
 		}
-		if tc.wantUpdatedConditionName == "" {
-			continue
-		}
-		pr, _ := client.ProvisioningRequestNoCache("ns", tc.wantUpdatedConditionName)
-		accepted := apimeta.FindStatusCondition(pr.Status.Conditions, v1.Accepted)
-		if accepted == nil || accepted.LastTransitionTime != metav1.NewTime(now) {
-			t.Errorf("%s: injector.Process hasn't update accepted condition for ProvisioningRequest %s", tc.name, tc.wantUpdatedConditionName)
-		}
 	}
 
 }
diff --git a/cluster-autoscaler/provisioningrequest/besteffortatomic/provisioning_class.go b/cluster-autoscaler/provisioningrequest/besteffortatomic/provisioning_class.go
index 122312b622cc..f21e507c0eeb 100644
--- a/cluster-autoscaler/provisioningrequest/besteffortatomic/provisioning_class.go
+++ b/cluster-autoscaler/provisioningrequest/besteffortatomic/provisioning_class.go
@@ -94,6 +94,7 @@ func (o *bestEffortAtomicProvClass) Provision(
 
 	// For provisioning requests, unschedulablePods are actually all injected pods. Some may even be schedulable!
 	actuallyUnschedulablePods, err := o.filterOutSchedulable(unschedulablePods)
+	conditions.AddOrUpdateCondition(pr, v1.Accepted, metav1.ConditionTrue, conditions.AcceptedReason, conditions.AcceptedMsg, metav1.Now())
 	if err != nil {
 		conditions.AddOrUpdateCondition(pr, v1.Provisioned, metav1.ConditionFalse, conditions.FailedToCheckCapacityReason, conditions.FailedToCheckCapacityMsg, metav1.Now())
 		if _, updateErr := o.client.UpdateProvisioningRequest(pr.ProvisioningRequest); updateErr != nil {
diff --git a/cluster-autoscaler/provisioningrequest/checkcapacity/provisioningclass.go b/cluster-autoscaler/provisioningrequest/checkcapacity/provisioningclass.go
index 1a3b45cbbaf5..244c9a0bd280 100644
--- a/cluster-autoscaler/provisioningrequest/checkcapacity/provisioningclass.go
+++ b/cluster-autoscaler/provisioningrequest/checkcapacity/provisioningclass.go
@@ -138,6 +138,7 @@ func (o *checkCapacityProvClass) checkcapacity(unschedulablePods []*apiv1.Pod, p
 	var capacityAvailable bool
 	err, cleanupErr := clustersnapshot.WithForkedSnapshot(o.context.ClusterSnapshot, func() (bool, error) {
 		st, _, err := o.schedulingSimulator.TrySchedulePods(o.context.ClusterSnapshot, unschedulablePods, scheduling.ScheduleAnywhere, true)
+		conditions.AddOrUpdateCondition(provReq, v1.Accepted, metav1.ConditionTrue, conditions.AcceptedReason, conditions.AcceptedMsg, metav1.Now())
 		if len(st) < len(unschedulablePods) || err != nil {
 			conditions.AddOrUpdateCondition(provReq, v1.Provisioned, metav1.ConditionFalse, conditions.CapacityIsNotFoundReason, "Capacity is not found, CA will try to find it later.", metav1.Now())
 			capacityAvailable = false