diff --git a/vllm/v1/worker/gpu_worker.py b/vllm/v1/worker/gpu_worker.py
index 9082bbfd8f8e6..8c75e8914857b 100644
--- a/vllm/v1/worker/gpu_worker.py
+++ b/vllm/v1/worker/gpu_worker.py
@@ -155,14 +155,6 @@ class Worker(WorkerBase):
 
     def init_device(self):
         if self.device_config.device.type == "cuda":
-            # torch.distributed.all_reduce does not free the input tensor until
-            # the synchronization point. This causes the memory usage to grow
-            # as the number of all_reduce calls increases. This env var disables
-            # this behavior.
-            # Related issue:
-            # https://discuss.pytorch.org/t/cuda-allocation-lifetime-for-inputs-to-distributed-all-reduce/191573
-            os.environ["TORCH_NCCL_AVOID_RECORD_STREAMS"] = "1"
-
             # This env var set by Ray causes exceptions with graph building.
             os.environ.pop("NCCL_ASYNC_ERROR_HANDLING", None)
             self.device = torch.device(f"cuda:{self.local_rank}")