diff --git a/weed/mq/kafka/integration/broker_client_subscribe.go b/weed/mq/kafka/integration/broker_client_subscribe.go
index 721b71b22..31f6efaa6 100644
--- a/weed/mq/kafka/integration/broker_client_subscribe.go
+++ b/weed/mq/kafka/integration/broker_client_subscribe.go
@@ -555,7 +555,9 @@ func (bc *BrokerClient) ReadRecords(ctx context.Context, session *BrokerSubscrib
 		}
 
 		readStart := time.Now()
-		ctx2, cancel2 := context.WithTimeout(context.Background(), currentTimeout)
+		// CRITICAL: Use parent context (ctx) to respect client's MaxWaitTime deadline
+		// The per-record timeout is combined with the overall fetch deadline
+		ctx2, cancel2 := context.WithTimeout(ctx, currentTimeout)
 		recvChan2 := make(chan recvResult, 1)
 
 		go func() {
diff --git a/weed/mq/kafka/protocol/fetch.go b/weed/mq/kafka/protocol/fetch.go
index 06dba826f..611f2fbc6 100644
--- a/weed/mq/kafka/protocol/fetch.go
+++ b/weed/mq/kafka/protocol/fetch.go
@@ -244,9 +244,8 @@ func (h *Handler) handleFetch(ctx context.Context, correlationID uint32, apiVers
 	// Phase 2: Wait for all results with adequate timeout for CI environments
 	// CRITICAL: We MUST return a result for every requested partition or Sarama will error
 	results := make([]*partitionFetchResult, len(pending))
-	// Deadline must be longer than subscriber's record batching timeout (1s) to avoid premature timeout
-	// Add 500ms buffer to account for network and processing overhead
-	deadline := time.After(1500 * time.Millisecond)
+	// Use the client's requested MaxWaitTime
+	deadline := time.After(time.Duration(maxWaitMs) * time.Millisecond)
 
 	// Collect results one by one with shared deadline
 	for i, pf := range pending {