diff --git a/sentry_sdk/_types.py b/sentry_sdk/_types.py
index 7043bbc2ee..ecb8abcd10 100644
--- a/sentry_sdk/_types.py
+++ b/sentry_sdk/_types.py
@@ -359,3 +359,7 @@ class SDKInfo(TypedDict):
     )
 
     HttpStatusCodeRange = Union[int, Container[int]]
+
+    class TextPart(TypedDict):
+        type: Literal["text"]
+        content: str
diff --git a/sentry_sdk/consts.py b/sentry_sdk/consts.py
index 93fca6ba3e..4b61a317fb 100644
--- a/sentry_sdk/consts.py
+++ b/sentry_sdk/consts.py
@@ -542,6 +542,12 @@ class SPANDATA:
     Example: 2048
     """
 
+    GEN_AI_SYSTEM_INSTRUCTIONS = "gen_ai.system_instructions"
+    """
+    The system instructions passed to the model.
+    Example: [{"type": "text", "text": "You are a helpful assistant."},{"type": "text", "text": "Be concise and clear."}]
+    """
+
     GEN_AI_REQUEST_MESSAGES = "gen_ai.request.messages"
     """
     The messages passed to the model. The "content" can be a string or an array of objects.
diff --git a/sentry_sdk/integrations/langchain.py b/sentry_sdk/integrations/langchain.py
index 8a9bc5167c..1cccfb0fff 100644
--- a/sentry_sdk/integrations/langchain.py
+++ b/sentry_sdk/integrations/langchain.py
@@ -36,6 +36,7 @@
     from uuid import UUID
 
     from sentry_sdk.tracing import Span
+    from sentry_sdk._types import TextPart
 
 
 try:
@@ -189,6 +190,29 @@ def _get_current_agent() -> "Optional[str]":
     return None
 
 
+def _get_system_instructions(messages: "List[List[BaseMessage]]") -> "List[TextPart]":
+    system_instructions = []
+
+    for list_ in messages:
+        for message in list_:
+            if message.type == "system":
+                system_instructions.append(message)
+
+    return system_instructions
+
+
+def _transform_system_instructions(
+    system_instructions: "List[BaseMessage]",
+) -> "List[TextPart]":
+    return [
+        {
+            "type": "text",
+            "content": instruction.content,
+        }
+        for instruction in system_instructions
+    ]
+
+
 class LangchainIntegration(Integration):
     identifier = "langchain"
     origin = f"auto.ai.{identifier}"
@@ -430,9 +454,21 @@ def on_chat_model_start(
             _set_tools_on_span(span, all_params.get("tools"))
 
             if should_send_default_pii() and self.include_prompts:
+                system_instructions = _get_system_instructions(messages)
+                if len(system_instructions) > 0:
+                    set_data_normalized(
+                        span,
+                        SPANDATA.GEN_AI_SYSTEM_INSTRUCTIONS,
+                        _transform_system_instructions(system_instructions),
+                        unpack=False,
+                    )
+
                 normalized_messages = []
                 for list_ in messages:
                     for message in list_:
+                        if message.type == "system":
+                            continue
+
                         normalized_messages.append(
                             self._normalize_langchain_message(message)
                         )
diff --git a/tests/integrations/langchain/test_langchain.py b/tests/integrations/langchain/test_langchain.py
index 6f5f9f14a1..54c4664f3e 100644
--- a/tests/integrations/langchain/test_langchain.py
+++ b/tests/integrations/langchain/test_langchain.py
@@ -217,17 +217,21 @@ def test_langchain_agent(
         assert chat_spans[1]["data"]["gen_ai.usage.total_tokens"] == 117
 
     if send_default_pii and include_prompts:
-        assert (
-            "You are very powerful"
-            in chat_spans[0]["data"][SPANDATA.GEN_AI_REQUEST_MESSAGES]
-        )
+        assert [
+            {
+                "type": "text",
+                "content": "You are very powerful assistant, but don't know current events",
+            }
+        ] == json.loads(chat_spans[0]["data"][SPANDATA.GEN_AI_SYSTEM_INSTRUCTIONS])
         assert "5" in chat_spans[0]["data"][SPANDATA.GEN_AI_RESPONSE_TEXT]
         assert "word" in tool_exec_span["data"][SPANDATA.GEN_AI_TOOL_INPUT]
         assert 5 == int(tool_exec_span["data"][SPANDATA.GEN_AI_TOOL_OUTPUT])
-        assert (
-            "You are very powerful"
-            in chat_spans[1]["data"][SPANDATA.GEN_AI_REQUEST_MESSAGES]
-        )
+        assert [
+            {
+                "type": "text",
+                "content": "You are very powerful assistant, but don't know current events",
+            }
+        ] == json.loads(chat_spans[1]["data"][SPANDATA.GEN_AI_SYSTEM_INSTRUCTIONS])
         assert "5" in chat_spans[1]["data"][SPANDATA.GEN_AI_RESPONSE_TEXT]
 
         # Verify tool calls are recorded when PII is enabled
@@ -243,8 +247,10 @@ def test_langchain_agent(
             tool_call_str = str(tool_calls_data)
             assert "get_word_length" in tool_call_str
     else:
+        assert SPANDATA.GEN_AI_SYSTEM_INSTRUCTIONS not in chat_spans[0].get("data", {})
         assert SPANDATA.GEN_AI_REQUEST_MESSAGES not in chat_spans[0].get("data", {})
         assert SPANDATA.GEN_AI_RESPONSE_TEXT not in chat_spans[0].get("data", {})
+        assert SPANDATA.GEN_AI_SYSTEM_INSTRUCTIONS not in chat_spans[1].get("data", {})
         assert SPANDATA.GEN_AI_REQUEST_MESSAGES not in chat_spans[1].get("data", {})
         assert SPANDATA.GEN_AI_RESPONSE_TEXT not in chat_spans[1].get("data", {})
         assert SPANDATA.GEN_AI_TOOL_INPUT not in tool_exec_span.get("data", {})