diff --git a/basecamp_client.py b/basecamp_client.py
index 7e2b27b..e6f15cf 100644
--- a/basecamp_client.py
+++ b/basecamp_client.py
@@ -405,21 +405,46 @@ class BasecampClient:
             raise Exception(f"Failed to get schedule: {str(e)}")
 
     # Comments methods
-    def get_comments(self, project_id, recording_id):
+    def get_comments(self, project_id, recording_id, page=1):
         """
-        Get all comments for a recording (todos, message, etc.).
-        
+        Get comments for a recording (todos, message, etc.).
+
         Args:
+            project_id (int): Project/bucket ID.
             recording_id (int): ID of the recording (todos, message, etc.)
-            project_id (int): Project/bucket ID. If not provided, it will be extracted from the recording ID.
-            
+            page (int): Page number for pagination (default: 1).
+                        Basecamp uses geared pagination: page 1 has 15 results,
+                        page 2 has 30, page 3 has 50, page 4+ has 100.
+
         Returns:
-            list: Comments for the recording
+            dict: Contains 'comments' list and pagination metadata:
+                  - comments: list of comments
+                  - total_count: total number of comments (from X-Total-Count header)
+                  - next_page: next page number if available, None otherwise
         """
         endpoint = f"buckets/{project_id}/recordings/{recording_id}/comments.json"
-        response = self.get(endpoint)
+        response = self.get(endpoint, params={"page": page})
         if response.status_code == 200:
-            return response.json()
+            # Parse pagination headers
+            total_count = response.headers.get('X-Total-Count')
+            total_count = int(total_count) if total_count else None
+
+            # Parse Link header for next page
+            next_page = None
+            link_header = response.headers.get('Link', '')
+            if 'rel="next"' in link_header:
+                # Extract page number from Link header
+                # Format: <https://3.basecampapi.com/.../comments.json?page=2>; rel="next"
+                import re
+                match = re.search(r'page=(\d+).*rel="next"', link_header)
+                if match:
+                    next_page = int(match.group(1))
+
+            return {
+                "comments": response.json(),
+                "total_count": total_count,
+                "next_page": next_page
+            }
         else:
             raise Exception(f"Failed to get comments: {response.status_code} - {response.text}")
 
diff --git a/basecamp_fastmcp.py b/basecamp_fastmcp.py
index aeebe30..e7b51cf 100644
--- a/basecamp_fastmcp.py
+++ b/basecamp_fastmcp.py
@@ -500,23 +500,28 @@ async def global_search(query: str) -> Dict[str, Any]:
         }
 
 @mcp.tool()
-async def get_comments(recording_id: str, project_id: str) -> Dict[str, Any]:
+async def get_comments(recording_id: str, project_id: str, page: int = 1) -> Dict[str, Any]:
     """Get comments for a Basecamp item.
-    
+
     Args:
         recording_id: The item ID
         project_id: The project ID
+        page: Page number for pagination (default: 1). Basecamp uses geared pagination:
+              page 1 has 15 results, page 2 has 30, page 3 has 50, page 4+ has 100.
     """
     client = _get_basecamp_client()
     if not client:
         return _get_auth_error_response()
-    
+
     try:
-        comments = await _run_sync(client.get_comments, project_id, recording_id)
+        result = await _run_sync(client.get_comments, project_id, recording_id, page)
         return {
             "status": "success",
-            "comments": comments,
-            "count": len(comments)
+            "comments": result["comments"],
+            "count": len(result["comments"]),
+            "page": page,
+            "total_count": result["total_count"],
+            "next_page": result["next_page"]
         }
     except Exception as e:
         logger.error(f"Error getting comments: {e}")
diff --git a/mcp_server_cli.py b/mcp_server_cli.py
index e1a59a2..27ae13a 100755
--- a/mcp_server_cli.py
+++ b/mcp_server_cli.py
@@ -193,7 +193,8 @@ class MCPServer:
                     "type": "object",
                     "properties": {
                         "recording_id": {"type": "string", "description": "The item ID"},
-                        "project_id": {"type": "string", "description": "The project ID"}
+                        "project_id": {"type": "string", "description": "The project ID"},
+                        "page": {"type": "integer", "description": "Page number for pagination (default: 1). Basecamp uses geared pagination: page 1 has 15 results, page 2 has 30, page 3 has 50, page 4+ has 100.", "default": 1}
                     },
                     "required": ["recording_id", "project_id"]
                 }
@@ -1025,11 +1026,15 @@ class MCPServer:
             elif tool_name == "get_comments":
                 recording_id = arguments.get("recording_id")
                 project_id = arguments.get("project_id")
-                comments = client.get_comments(project_id, recording_id)
+                page = arguments.get("page", 1)
+                result = client.get_comments(project_id, recording_id, page)
                 return {
                     "status": "success",
-                    "comments": comments,
-                    "count": len(comments)
+                    "comments": result["comments"],
+                    "count": len(result["comments"]),
+                    "page": page,
+                    "total_count": result["total_count"],
+                    "next_page": result["next_page"]
                 }
 
             elif tool_name == "create_comment":
diff --git a/search_utils.py b/search_utils.py
index b980d26..6906c26 100644
--- a/search_utils.py
+++ b/search_utils.py
@@ -467,7 +467,8 @@ class BasecampSearch:
         try:
             # If both recording_id and bucket_id are provided, get comments for that specific recording
             if recording_id and bucket_id:
-                comments = self.client.get_comments(recording_id, bucket_id)
+                result = self.client.get_comments(recording_id, bucket_id)
+                comments = result["comments"]
             # Otherwise we can't search across all comments as there's no endpoint for that
             else:
                 logger.warning("Cannot search all comments across Basecamp - both recording_id and bucket_id are required")