diff --git a/vllm/distributed/kv_transfer/kv_connector/utils.py b/vllm/distributed/kv_transfer/kv_connector/utils.py
index 1f889c6c8..72980a85a 100644
--- a/vllm/distributed/kv_transfer/kv_connector/utils.py
+++ b/vllm/distributed/kv_transfer/kv_connector/utils.py
@@ -457,9 +457,11 @@ class TpKVTopology:
         """
         Whether the KV cache is replicated across TP workers due to the
         number of TP workers being greater than the number of KV heads.
+        When they are equal, each TP rank still owns one distinct KV head,
+        so this is not considered replication.
         """
         tp_size = self.remote_tp_size[engine_id]
-        return tp_size // self.total_num_kv_heads >= 1
+        return tp_size > self.total_num_kv_heads
 
     def replicates_kv_cache(self, remote_engine_id: EngineId) -> bool:
         # MLA is always replicated as the hidden dim can't be split.