diff --git a/internal/pkg/object/command/ecs/ecs.go b/internal/pkg/object/command/ecs/ecs.go
index d53e202..eec206a 100644
--- a/internal/pkg/object/command/ecs/ecs.go
+++ b/internal/pkg/object/command/ecs/ecs.go
@@ -20,6 +20,7 @@ import (
 	"github.com/patterninc/heimdall/pkg/object/cluster"
 	"github.com/patterninc/heimdall/pkg/object/job"
 	"github.com/patterninc/heimdall/pkg/plugin"
+	"github.com/pkg/errors"
 )
 
 // ECS command context structure
@@ -676,6 +677,7 @@ func (execCtx *executionContext) retrieveLogs(ctx context.Context) error {
 
 }
 func (e *commandContext) Cleanup(ctx context.Context, jobID string, c *cluster.Cluster) error {
+
 	cleanupMethod.CountRequest()
 	// Resolve cluster context to get cluster name
 	clusterContext := &clusterContext{}
@@ -691,8 +693,10 @@ func (e *commandContext) Cleanup(ctx context.Context, jobID string, c *cluster.C
 	ecsClient := ecs.NewFromConfig(cfg)
 
 	// List all tasks started by this job
-	var allTaskARNs []string
-	for taskNum := 0; taskNum < e.TaskCount; taskNum++ {
+	maxTaskCount := clusterContext.MaxTaskCount
+
+	taskARNs := make([]string, 0)
+	for taskNum := 0; taskNum < maxTaskCount; taskNum++ {
 		startedByValue := fmt.Sprintf("%s%s-%d", startedByPrefix, jobID, taskNum)
 
 		listTasksOutput, err := ecsClient.ListTasks(ctx, &ecs.ListTasksInput{
@@ -703,44 +707,32 @@ func (e *commandContext) Cleanup(ctx context.Context, jobID string, c *cluster.C
 			cleanupMethod.CountError("list_tasks")
 			return err
 		}
-		allTaskARNs = append(allTaskARNs, listTasksOutput.TaskArns...)
+
+		taskARNs = append(taskARNs, listTasksOutput.TaskArns...)
+
+		time.Sleep(100 * time.Millisecond) // prevent API throttling
 	}
 
-	if len(allTaskARNs) == 0 {
+	if len(taskARNs) == 0 {
 		// No tasks found, nothing to clean up
 		cleanupMethod.CountSuccess("no_tasks_found")
 		return nil
 	}
 
-	// Bulk describe all tasks to check their LastStatus
-	describeOutput, err := ecsClient.DescribeTasks(ctx, &ecs.DescribeTasksInput{
-		Cluster: aws.String(clusterContext.ClusterName),
-		Tasks:   allTaskARNs,
-	})
-	if err != nil {
-		cleanupMethod.CountError("describe_tasks")
-		return err
-	}
-
-	// Stop all tasks where LastStatus != STOPPED or SUCCEEDED
-	for _, task := range describeOutput.Tasks {
-		// Skip tasks that are already stopped
-		if aws.ToString(task.LastStatus) == "STOPPED" || aws.ToString(task.LastStatus) == "SUCCEEDED" {
-			continue
-		}
-
+	// Stop all tasks we found. StopTask is safe to call even if the task is already stopping/stopped.
+	for _, taskARN := range taskARNs {
 		stopTaskInput := &ecs.StopTaskInput{
 			Cluster: aws.String(clusterContext.ClusterName),
-			Task:    task.TaskArn,
+			Task:    aws.String(taskARN),
 			Reason:  aws.String(errJobTerminated),
 		}
-		_, err := ecsClient.StopTask(ctx, stopTaskInput)
-		if err != nil {
+		if _, err := ecsClient.StopTask(ctx, stopTaskInput); err != nil {
 			// Log error but continue stopping other tasks
-			cleanupMethod.LogAndCountError(err, fmt.Sprintf("failed to stop task %s", aws.ToString(task.TaskArn)))
-			continue
+			err = errors.Wrapf(err, "failed to stop task %s", taskARN)
+			cleanupMethod.LogAndCountError(err, "stop_task")
 		}
-		cleanupMethod.CountSuccess("stop_task")
+
+		time.Sleep(100 * time.Millisecond) // prevent API throttling
 	}
 	cleanupMethod.CountSuccess()
 	return nil