diff --git a/pandas/io/parquet.py b/pandas/io/parquet.py
index 56131d000b176..3b8077c531996 100644
--- a/pandas/io/parquet.py
+++ b/pandas/io/parquet.py
@@ -173,9 +173,25 @@ def write(
 
         table = self.api.Table.from_pandas(df, **from_pandas_kwargs)
 
+        filesystem = kwargs.pop("filesystem", None)
+        if (
+            isinstance(path, str)
+            and storage_options is None
+            and filesystem is None
+            and Version(self.api.__version__) >= Version("5.0.0")
+        ):
+            try:
+                from pyarrow.fs import FileSystem
+
+                filesystem, path = FileSystem.from_uri(path)
+            except Exception:
+                # fallback to use get_handle / fsspec for filesystems
+                # that pyarrow doesn't support
+                pass
+
         path_or_handle, handles, kwargs["filesystem"] = _get_path_or_handle(
             path,
-            kwargs.pop("filesystem", None),
+            filesystem,
             storage_options=storage_options,
             mode="wb",
             is_dir=partition_cols is not None,
@@ -470,6 +486,12 @@ def read_parquet(
 
         .. versionadded:: 1.3.0
 
+        When using the 'pyarrow' engine, no storage options are provided
+        and a filesystem is implemented by both ``pyarrow.fs`` and ``fsspec``
+        (e.g. "s3://"), then the ``pyarrow.fs`` filesystem is preferred.
+        Provide the instantiated fsspec filesystem using the ``filesystem``
+        keyword if you wish to use its implementation.
+
     use_nullable_dtypes : bool, default False
         If True, use dtypes that use ``pd.NA`` as missing value indicator
         for the resulting DataFrame. (only applicable for the ``pyarrow``