diff --git a/pyathena/polars/result_set.py b/pyathena/polars/result_set.py
index bd1e10505..3eab3059b 100644
--- a/pyathena/polars/result_set.py
+++ b/pyathena/polars/result_set.py
@@ -456,9 +456,7 @@ def _get_parquet_storage_options(connection: Connection[Any]) -> dict[str, Any]:
         fsspec. The credentials are frozen when the options are created.
 
         Args:
-            connection: The connection whose session and region the reads use. The
-                chunk readers give the connection that they took when they were
-                created, as ``close()`` releases the result set's connection.
+            connection: The connection whose session and region the reads use.
 
         Returns:
             Dictionary with AWS credentials and region for S3 access.
@@ -606,9 +604,12 @@ def _create_dataframe_iterator(self) -> PolarsDataFrameIterator:
         For unload queries, it replaces the metadata with the schema of the Parquet
         files, as ``_as_polars()`` does, before the chunks are read.
 
-        The chunk readers take the state that they need, such as the output location,
-        the column names, and the connection, when they are created. The credentials
-        are frozen at the first read.
+        The result set calls this method when it is created, in ``execute()`` or in the
+        future that ``AsyncPolarsCursor.execute()`` returns. The requests that prepare
+        the reads are made then, before the first fetch, and raise their errors there:
+        the size check of a CSV result file, and the data manifest and the Parquet
+        schema of an UNLOAD result. The schema sets the metadata and is passed to
+        ``_iter_parquet_chunks()``. The chunks themselves are read lazily.
 
         Returns:
             PolarsDataFrameIterator that reads each chunk lazily.