diff --git a/vllm/entrypoints/cli/serve.py b/vllm/entrypoints/cli/serve.py
index 2ca31510208d..23ac468ffa11 100644
--- a/vllm/entrypoints/cli/serve.py
+++ b/vllm/entrypoints/cli/serve.py
@@ -119,6 +119,10 @@ def run_headless(args: argparse.Namespace):
     if not envs.VLLM_USE_V1:
         raise ValueError("Headless mode is only supported for V1")
 
+    if engine_args.data_parallel_rank is not None:
+        raise ValueError("data_parallel_rank is not applicable in "
+                         "headless mode")
+
     parallel_config = vllm_config.parallel_config
     local_engine_count = parallel_config.data_parallel_size_local
 
@@ -126,10 +130,6 @@ def run_headless(args: argparse.Namespace):
         raise ValueError("data_parallel_size_local must be > 0 in "
                          "headless mode")
 
-    if parallel_config.data_parallel_rank is not None:
-        raise ValueError("data_parallel_rank is not applicable in "
-                         "headless mode")
-
     host = parallel_config.data_parallel_master_ip
     port = engine_args.data_parallel_rpc_port  # add to config too
     handshake_address = get_tcp_uri(host, port)