diff --git a/litellm/llms/vertex_ai.py b/litellm/llms/vertex_ai.py
index b52e8689f10..4452ae43a0d 100644
--- a/litellm/llms/vertex_ai.py
+++ b/litellm/llms/vertex_ai.py
@@ -3,7 +3,7 @@ import json
 from enum import Enum
 import requests  # type: ignore
 import time
-from typing import Callable, Optional, Union, List, Literal
+from typing import Callable, Optional, Union, List, Literal, Tuple
 from litellm.utils import ModelResponse, Usage, CustomStreamWrapper, map_finish_reason
 import litellm, uuid
 import httpx, inspect  # type: ignore
@@ -336,14 +336,39 @@ def _process_gemini_image(image_url: str) -> PartType:
         raise e
 
 
-def _gemini_convert_messages_with_history(messages: list) -> List[ContentType]:
+def _extract_system_prompt_from_messages(messages: list) -> Optional[ContentType]:
+    # Separate system prompt from rest of message
+    system_prompt_indices = []
+    _parts: List[PartType] = []
+    for idx, message in enumerate(messages):
+        if message["role"] == "system":
+            _part = PartType(text=message["content"])
+            _parts.append(_part)
+            system_prompt_indices.append(idx)
+    if len(system_prompt_indices) > 0:
+        for idx in reversed(system_prompt_indices):
+            messages.pop(idx)
+    if len(_parts) > 0:
+        return ContentType(parts=_parts)
+    return None
+
+
+def _gemini_convert_messages_with_history(
+    messages: list,
+) -> Tuple[Optional[ContentType], List[ContentType]]:
     """
     Converts given messages from OpenAI format to Gemini format
 
     - Parts must be iterable
     - Roles must alternate b/w 'user' and 'model' (same as anthropic -> merge consecutive roles)
     - Please ensure that function response turn comes immediately after a function call turn
+
+    Returns:
+    - Tuple[Optional[system_instructions], messages]
     """
+
+    system_instructions = _extract_system_prompt_from_messages(messages=messages)
+
     user_message_types = {"user", "system"}
     contents: List[ContentType] = []
 
@@ -404,7 +429,7 @@ def _gemini_convert_messages_with_history(messages: list) -> List[ContentType]:
                 )
             )
 
-    return contents
+    return system_instructions, contents
 
 
 def _gemini_vision_convert_messages(messages: list):
@@ -699,7 +724,9 @@ def completion(
             print_verbose("\nMaking VertexAI Gemini Pro / Pro Vision Call")
             print_verbose(f"\nProcessing input messages = {messages}")
             tools = optional_params.pop("tools", None)
-            content = _gemini_convert_messages_with_history(messages=messages)
+            system_instruction, content = _gemini_convert_messages_with_history(
+                messages=messages
+            )
             stream = optional_params.pop("stream", False)
             if stream == True:
                 request_str += f"response = llm_model.generate_content({content}, generation_config=GenerationConfig(**{optional_params}), safety_settings={safety_settings}, stream={stream})\n"
@@ -736,6 +763,7 @@ def completion(
             ## LLM Call
             response = llm_model.generate_content(
                 contents=content,
+                system_instruction=system_instruction,
                 generation_config=optional_params,
                 safety_settings=safety_settings,
                 tools=tools,
diff --git a/litellm/tests/test_amazing_vertex_completion.py b/litellm/tests/test_amazing_vertex_completion.py
index d7b2dc2d403..4df39fec087 100644
--- a/litellm/tests/test_amazing_vertex_completion.py
+++ b/litellm/tests/test_amazing_vertex_completion.py
@@ -676,7 +676,7 @@ async def test_gemini_pro_function_calling(sync_mode):
 # gemini_pro_function_calling()
 
 
-@pytest.mark.parametrize("sync_mode", [False, True])
+@pytest.mark.parametrize("sync_mode", [True])
 @pytest.mark.asyncio
 async def test_gemini_pro_function_calling_streaming(sync_mode):
     load_vertex_ai_credentials()