diff --git a/vllm_ascend/worker.py b/vllm_ascend/worker.py
index 27930d4..d1e0765 100644
--- a/vllm_ascend/worker.py
+++ b/vllm_ascend/worker.py
@@ -84,9 +84,6 @@ class NPUWorker(LocalOrDistributedWorkerBase):
         self.distributed_init_method = distributed_init_method
         self.is_driver_worker = is_driver_worker
 
-        if is_driver_worker:
-            assert rank % self.parallel_config.tensor_parallel_size == 0, \
-                   "Driver worker should be rank 0 of tensor parallel group."
         if self.model_config.trust_remote_code:
             # note: lazy import to avoid importing torch before initializing
             from vllm.utils import init_cached_hf_modules