diff --git a/doc/source/whatsnew/v1.1.0.rst b/doc/source/whatsnew/v1.1.0.rst
index f216418c3a8b0..4c8bf62490d82 100644
--- a/doc/source/whatsnew/v1.1.0.rst
+++ b/doc/source/whatsnew/v1.1.0.rst
@@ -991,6 +991,7 @@ I/O
 - Bug in :meth:`~SQLDatabase.execute` was raising a ``ProgrammingError`` for some DB-API drivers when the SQL statement contained the `%` character and no parameters were present (:issue:`34211`)
 - Bug in :meth:`~pandas.io.stata.StataReader` which resulted in categorical variables with difference dtypes when reading data using an iterator. (:issue:`31544`)
 - :meth:`HDFStore.keys` has now an optional `include` parameter that allows the retrieval of all native HDF5 table names (:issue:`29916`)
+- `pandas.io.parquet.PyArrowImpl` now infers `filesystem` using the provided `path` if `filesystem` is not provided via `kwargs`. (:issue:`34841`)
 
 Plotting
 ^^^^^^^^
diff --git a/pandas/io/parquet.py b/pandas/io/parquet.py
index de9a14c82b3cb..530bb6e326e94 100644
--- a/pandas/io/parquet.py
+++ b/pandas/io/parquet.py
@@ -104,6 +104,9 @@ def write(
             from_pandas_kwargs["preserve_index"] = index
 
         table = self.api.Table.from_pandas(df, **from_pandas_kwargs)
+
+        fs = kwargs.pop("filesystem", get_fs_for_path(path))
+
         # write_to_dataset does not support a file-like object when
         # a directory path is used, so just pass the path string.
         if partition_cols is not None:
@@ -111,12 +114,13 @@ def write(
                 table,
                 path,
                 compression=compression,
+                filesystem=fs,
                 partition_cols=partition_cols,
                 **kwargs,
             )
         else:
             self.api.parquet.write_table(
-                table, file_obj_or_path, compression=compression, **kwargs
+                table, file_obj_or_path, compression=compression, **kwargs,
             )
         if should_close:
             file_obj_or_path.close()
diff --git a/pandas/tests/io/test_parquet.py b/pandas/tests/io/test_parquet.py
index efd34c58d7d19..b6ed8e8b3652a 100644
--- a/pandas/tests/io/test_parquet.py
+++ b/pandas/tests/io/test_parquet.py
@@ -568,6 +568,24 @@ def test_s3_roundtrip_for_dir(self, df_compat, s3_resource, pa, partition_col):
             repeat=1,
         )
 
+    @td.skip_if_no("s3fs")
+    @pytest.mark.parametrize("partition_col", [["A"], []])
+    def test_s3_roundtrip_for_dir_infer_fs(
+        self, df_compat, s3_resource, pa, partition_col
+    ):
+        expected_df = df_compat.copy()
+        if partition_col:
+            expected_df[partition_col] = expected_df[partition_col].astype("category")
+        check_round_trip(
+            df_compat,
+            pa,
+            expected=expected_df,
+            path="s3://pandas-test/parquet_dir",
+            write_kwargs={"partition_cols": partition_col, "compression": None},
+            check_like=True,
+            repeat=1,
+        )
+
     @tm.network
     @td.skip_if_no("pyarrow")
     def test_parquet_read_from_url(self, df_compat):