diff --git a/litellm/main.py b/litellm/main.py
index 7775922ea44..7db4e5648e3 100644
--- a/litellm/main.py
+++ b/litellm/main.py
@@ -6062,7 +6062,7 @@ def completion(
             custom_llm_provider=custom_llm_provider,
             model_id=optional_params.get("model_id"),
         )
-        if mock_response or mock_tool_calls or mock_timeout:
+        if mock_response is not None or mock_tool_calls or mock_timeout:
             kwargs.pop("mock_timeout", None)  # remove for any fallbacks triggered
             return mock_completion(
                 model,
diff --git a/tests/unit/test_empty_mock_response.py b/tests/unit/test_empty_mock_response.py
new file mode 100644
index 00000000000..dbae14f344d
--- /dev/null
+++ b/tests/unit/test_empty_mock_response.py
@@ -0,0 +1,46 @@
+from unittest.mock import patch
+
+import pytest
+
+import litellm
+
+
+@pytest.mark.parametrize("mock_response", ["", "Hello"])
+@pytest.mark.parametrize("stream", [False, True])
+def test_completion_mock_response(mock_response: str, stream: bool) -> None:
+    with patch("httpx.Client.send", side_effect=AssertionError("Unexpected HTTP request")):
+        response = litellm.completion(
+            model="openai/gpt-4o-mini",
+            messages=[{"role": "user", "content": "hi"}],
+            mock_response=mock_response,
+            stream=stream,
+        )
+        if stream:
+            content = "".join(chunk.choices[0].delta.content or "" for chunk in response)
+        else:
+            content = response.choices[0].message.content
+    assert content == mock_response
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize("mock_response", ["", "Hello"])
+@pytest.mark.parametrize("stream", [False, True])
+async def test_acompletion_mock_response(mock_response: str, stream: bool) -> None:
+    with (
+        patch(
+            "httpx.AsyncClient.send",
+            side_effect=AssertionError("Unexpected HTTP request"),
+        ),
+        patch("httpx.Client.send", side_effect=AssertionError("Unexpected HTTP request")),
+    ):
+        response = await litellm.acompletion(
+            model="openai/gpt-4o-mini",
+            messages=[{"role": "user", "content": "hi"}],
+            mock_response=mock_response,
+            stream=stream,
+        )
+        if stream:
+            content = "".join([chunk.choices[0].delta.content or "" async for chunk in response])
+        else:
+            content = response.choices[0].message.content
+    assert content == mock_response