diff --git a/src/strands/agent/conversation_manager/sliding_window_conversation_manager.py b/src/strands/agent/conversation_manager/sliding_window_conversation_manager.py
index f4c75cf1a..caaab4676 100644
--- a/src/strands/agent/conversation_manager/sliding_window_conversation_manager.py
+++ b/src/strands/agent/conversation_manager/sliding_window_conversation_manager.py
@@ -58,13 +58,8 @@ def __init__(self, window_size: int = 40, should_truncate_results: bool = True):
     def apply_management(self, agent: "Agent", **kwargs: Any) -> None:
         """Apply the sliding window to the agent's messages array to maintain a manageable history size.
 
-        This method is called after every event loop cycle, as the messages array may have been modified with tool
-        results and assistant responses. It first removes any dangling messages that might create an invalid
-        conversation state, then applies the sliding window if the message count exceeds the window size.
-
-        Special handling is implemented to ensure we don't leave a user message with toolResult
-        as the first message in the array. It also ensures that all toolUse blocks have corresponding toolResult
-        blocks to maintain conversation coherence.
+        This method is called after every event loop cycle to apply a sliding window if the message count
+        exceeds the window size.
 
         Args:
             agent: The agent whose messages will be managed.
@@ -72,7 +67,6 @@ def apply_management(self, agent: "Agent", **kwargs: Any) -> None:
             **kwargs: Additional keyword arguments for future extensibility.
         """
         messages = agent.messages
-        self._remove_dangling_messages(messages)
 
         if len(messages) <= self.window_size:
             logger.debug(
@@ -81,37 +75,6 @@ def apply_management(self, agent: "Agent", **kwargs: Any) -> None:
             return
         self.reduce_context(agent)
 
-    def _remove_dangling_messages(self, messages: Messages) -> None:
-        """Remove dangling messages that would create an invalid conversation state.
-
-        After the event loop cycle is executed, we expect the messages array to end with either an assistant tool use
-        request followed by the pairing user tool result or an assistant response with no tool use request. If the
-        event loop cycle fails, we may end up in an invalid message state, and so this method will remove problematic
-        messages from the end of the array.
-
-        This method handles two specific cases:
-
-        - User with no tool result: Indicates that event loop failed to generate an assistant tool use request
-        - Assistant with tool use request: Indicates that event loop failed to generate a pairing user tool result
-
-        Args:
-            messages: The messages to clean up.
-                This list is modified in-place.
-        """
-        # remove any dangling user messages with no ToolResult
-        if len(messages) > 0 and is_user_message(messages[-1]):
-            if not any("toolResult" in content for content in messages[-1]["content"]):
-                messages.pop()
-
-        # remove any dangling assistant messages with ToolUse
-        if len(messages) > 0 and is_assistant_message(messages[-1]):
-            if any("toolUse" in content for content in messages[-1]["content"]):
-                messages.pop()
-                # remove remaining dangling user messages with no ToolResult after we popped off an assistant message
-                if len(messages) > 0 and is_user_message(messages[-1]):
-                    if not any("toolResult" in content for content in messages[-1]["content"]):
-                        messages.pop()
-
     def reduce_context(self, agent: "Agent", e: Optional[Exception] = None, **kwargs: Any) -> None:
         """Trim the oldest messages to reduce the conversation context size.
 
diff --git a/tests/strands/agent/test_conversation_manager.py b/tests/strands/agent/test_conversation_manager.py
index 7d43199e8..db2e2cfb3 100644
--- a/tests/strands/agent/test_conversation_manager.py
+++ b/tests/strands/agent/test_conversation_manager.py
@@ -58,21 +58,21 @@ def conversation_manager(request):
                 {"role": "user", "content": [{"toolResult": {"toolUseId": "123", "content": [], "status": "success"}}]},
             ],
         ),
-        # 2 - Remove dangling user message with no tool result
+        # 2 - Keep user message
         (
             {"window_size": 2},
             [
                 {"role": "user", "content": [{"text": "Hello"}]},
             ],
-            [],
+            [{"role": "user", "content": [{"text": "Hello"}]}],
         ),
-        # 3 - Remove dangling assistant message with tool use
+        # 3 - Keep dangling assistant message with tool use
         (
             {"window_size": 3},
             [
                 {"role": "assistant", "content": [{"toolUse": {"toolUseId": "123", "name": "tool1", "input": {}}}]},
             ],
-            [],
+            [{"role": "assistant", "content": [{"toolUse": {"toolUseId": "123", "name": "tool1", "input": {}}}]}],
         ),
         # 4 - Remove dangling assistant message with tool use - User tool result remains
         (
@@ -83,6 +83,7 @@ def conversation_manager(request):
             ],
             [
                 {"role": "user", "content": [{"toolResult": {"toolUseId": "123", "content": [], "status": "success"}}]},
+                {"role": "assistant", "content": [{"toolUse": {"toolUseId": "123", "name": "tool1", "input": {}}}]},
             ],
         ),
         # 5 - Remove dangling assistant message with tool use and user message without tool result
@@ -95,8 +96,9 @@ def conversation_manager(request):
                 {"role": "assistant", "content": [{"toolUse": {"toolUseId": "123", "name": "tool1", "input": {}}}]},
             ],
             [
-                {"role": "user", "content": [{"text": "First"}]},
                 {"role": "assistant", "content": [{"text": "First response"}]},
+                {"role": "user", "content": [{"text": "Use a tool"}]},
+                {"role": "assistant", "content": [{"toolUse": {"toolUseId": "123", "name": "tool1", "input": {}}}]},
             ],
         ),
         # 6 - Message count above max window size - Basic drop