diff --git a/include/cutlass/epilogue/threadblock/epilogue_with_broadcast.h b/include/cutlass/epilogue/threadblock/epilogue_with_broadcast.h
index 7a97e0cb32..6b7351cfd1 100644
--- a/include/cutlass/epilogue/threadblock/epilogue_with_broadcast.h
+++ b/include/cutlass/epilogue/threadblock/epilogue_with_broadcast.h
@@ -946,13 +946,13 @@ class EpilogueWithBroadcast<
       //
 
       if (OutputOp::kStoreZ) {
+        destination_iterator += reduce_fragment_idx;
         destination_iterator.store(frag_Z);
-        ++destination_iterator;
       }
 
       if (OutputOp::kStoreT) {
+        tensor_iterator += reduce_fragment_idx;
         tensor_iterator.store(frag_T);
-        ++tensor_iterator;
       }
     }
 };
@@ -1698,13 +1698,13 @@ class EpilogueWithBroadcast<
       //
 
       if (OutputOp::kStoreZ) {
+        destination_iterator += reduce_fragment_idx;
         destination_iterator.store(frag_Z);
-        ++destination_iterator;
       }
 
       if (OutputOp::kStoreT) {
+        tensor_iterator += reduce_fragment_idx;
         tensor_iterator.store(frag_T);
-        ++tensor_iterator;
       }
     }
 };