diff --git a/pyarrow-stubs/_dataset.pyi b/pyarrow-stubs/_dataset.pyi
index d91a48d..fef3cbe 100644
--- a/pyarrow-stubs/_dataset.pyi
+++ b/pyarrow-stubs/_dataset.pyi
@@ -13,7 +13,7 @@ from typing import (
 )
 
 from . import _csv, _json, _parquet, lib
-from ._fs import FileSelector, FileSystem
+from ._fs import FileSelector, FileSystem, SupportedFileSystem
 from ._stubs_typing import Indices, JoinType, Order
 from .acero import ExecNodeOptions
 from .compute import Expression
@@ -129,7 +129,7 @@ class FileSystemDataset(Dataset):
         fragments: list[Fragment],
         schema: lib.Schema,
         format: FileFormat,
-        filesystem: FileSystem | None = None,
+        filesystem: SupportedFileSystem | None = None,
         root_partition: Expression | None = None,
     ) -> None: ...
     @classmethod
@@ -138,7 +138,7 @@ class FileSystemDataset(Dataset):
         paths: list[str],
         schema: lib.Schema | None = None,
         format: FileFormat | None = None,
-        filesystem: FileSystem | None = None,
+        filesystem: SupportedFileSystem | None = None,
         partitions: list[Expression] | None = None,
         root_partition: Expression | None = None,
     ) -> FileSystemDataset: ...
@@ -157,12 +157,12 @@ class FileWriteOptions(lib._Weakrefable):
 
 class FileFormat(lib._Weakrefable):
     def inspect(
-        self, file: str | Path | IO, filesystem: FileSystem | None = None
+        self, file: str | Path | IO, filesystem: SupportedFileSystem | None = None
     ) -> lib.Schema: ...
     def make_fragment(
         self,
         file: str | Path | IO,
-        filesystem: FileSystem | None = None,
+        filesystem: SupportedFileSystem | None = None,
         partition_expression: Expression | None = None,
         *,
         file_size: int | None = None,
@@ -402,7 +402,7 @@ class FileSystemFactoryOptions(lib._Weakrefable):
 class FileSystemDatasetFactory(DatasetFactory):
     def __init__(
         self,
-        filesystem: FileSystem,
+        filesystem: SupportedFileSystem,
         paths_or_selector: FileSelector,
         format: FileFormat,
         options: FileSystemFactoryOptions | None = None,
@@ -503,7 +503,7 @@ def _filesystemdataset_write(
     data: Scanner,
     base_dir: str | Path,
     basename_template: str,
-    filesystem: FileSystem,
+    filesystem: SupportedFileSystem,
     partitioning: Partitioning,
     file_options: FileWriteOptions,
     max_partitions: int,
diff --git a/pyarrow-stubs/_dataset_parquet.pyi b/pyarrow-stubs/_dataset_parquet.pyi
index ce1a840..f5b2c93 100644
--- a/pyarrow-stubs/_dataset_parquet.pyi
+++ b/pyarrow-stubs/_dataset_parquet.pyi
@@ -14,7 +14,7 @@ from ._dataset import (
     PartitioningFactory,
 )
 from ._dataset_parquet_encryption import ParquetDecryptionConfig
-from ._fs import FileSystem
+from ._fs import SupportedFileSystem
 from ._parquet import FileDecryptionProperties, FileMetaData
 from .lib import CacheOptions, Schema, _Weakrefable
 
@@ -36,7 +36,7 @@ class ParquetFileFormat(FileFormat):
     def make_fragment(
         self,
         file: IO | Path | str,
-        filesystem: FileSystem | None = None,
+        filesystem: SupportedFileSystem | None = None,
         partition_expression: Expression | None = None,
         row_groups: Iterable[int] | None = None,
         *,
@@ -118,7 +118,7 @@ class ParquetDatasetFactory(DatasetFactory):
     def __init__(
         self,
         metadata_path: str,
-        filesystem: FileSystem,
+        filesystem: SupportedFileSystem,
         format: FileFormat,
         options: ParquetFactoryOptions | None = None,
     ) -> None: ...
diff --git a/pyarrow-stubs/_fs.pyi b/pyarrow-stubs/_fs.pyi
index 3f87e3f..4725b20 100644
--- a/pyarrow-stubs/_fs.pyi
+++ b/pyarrow-stubs/_fs.pyi
@@ -2,10 +2,14 @@ import datetime as dt
 import enum
 
 from abc import ABC, abstractmethod
-from typing import Self, overload
+from typing import Self, TypeAlias, Union, overload
+
+from fsspec import AbstractFileSystem
 
 from .lib import NativeFile, _Weakrefable
 
+SupportedFileSystem: TypeAlias = Union[AbstractFileSystem, FileSystem]
+
 class FileType(enum.IntFlag):
     NotFound = enum.auto()
     Unknown = enum.auto()
diff --git a/pyarrow-stubs/dataset.pyi b/pyarrow-stubs/dataset.pyi
index 1fcc436..3473fe4 100644
--- a/pyarrow-stubs/dataset.pyi
+++ b/pyarrow-stubs/dataset.pyi
@@ -50,7 +50,7 @@ from pyarrow._dataset_parquet_encryption import (
 from pyarrow.compute import Expression, field, scalar
 from pyarrow.lib import Array, RecordBatch, RecordBatchReader, Schema, Table
 
-from ._fs import FileSystem
+from ._fs import SupportedFileSystem
 
 _orc_available: bool
 _parquet_available: bool
@@ -153,7 +153,7 @@ def partitioning(
 def parquet_dataset(
     metadata_path: str | Path,
     schema: Schema | None = None,
-    filesystem: FileSystem | None = None,
+    filesystem: SupportedFileSystem | None = None,
     format: ParquetFileFormat | None = None,
     partitioning: Partitioning | PartitioningFactory | None = None,
     partition_base_dir: str | None = None,
@@ -163,7 +163,7 @@ def dataset(
     source: str | list[str] | Path | list[Path],
     schema: Schema | None = None,
     format: FileFormat | _DatasetFormat | None = None,
-    filesystem: FileSystem | str | None = None,
+    filesystem: SupportedFileSystem | str | None = None,
     partitioning: Partitioning | PartitioningFactory | str | list[str] | None = None,
     partition_base_dir: str | None = None,
     exclude_invalid_files: bool | None = None,
@@ -174,7 +174,7 @@ def dataset(
     source: list[Dataset],
     schema: Schema | None = None,
     format: FileFormat | _DatasetFormat | None = None,
-    filesystem: FileSystem | str | None = None,
+    filesystem: SupportedFileSystem | str | None = None,
     partitioning: Partitioning | PartitioningFactory | str | list[str] | None = None,
     partition_base_dir: str | None = None,
     exclude_invalid_files: bool | None = None,
@@ -185,7 +185,7 @@ def dataset(
     source: Iterable[RecordBatch] | Iterable[Table] | RecordBatchReader,
     schema: Schema | None = None,
     format: FileFormat | _DatasetFormat | None = None,
-    filesystem: FileSystem | str | None = None,
+    filesystem: SupportedFileSystem | str | None = None,
     partitioning: Partitioning | PartitioningFactory | str | list[str] | None = None,
     partition_base_dir: str | None = None,
     exclude_invalid_files: bool | None = None,
@@ -200,7 +200,7 @@ def write_dataset(
     partitioning: Partitioning | list[str] | None = None,
     partitioning_flavor: str | None = None,
     schema: Schema | None = None,
-    filesystem: FileSystem | None = None,
+    filesystem: SupportedFileSystem | None = None,
     file_options: FileWriteOptions | None = None,
     use_threads: bool = True,
     max_partitions: int = 1024,
diff --git a/pyarrow-stubs/fs.pyi b/pyarrow-stubs/fs.pyi
index a185cb6..6bf7561 100644
--- a/pyarrow-stubs/fs.pyi
+++ b/pyarrow-stubs/fs.pyi
@@ -8,6 +8,7 @@ from pyarrow._fs import (  # noqa
     _MockFileSystem,
     FileSystemHandler,
     PyFileSystem,
+    SupportedFileSystem,
 )
 from pyarrow._azurefs import AzureFileSystem
 from pyarrow._hdfs import HadoopFileSystem
@@ -30,16 +31,16 @@ FileStats = FileInfo
 def copy_files(
     source: str,
     destination: str,
-    source_filesystem: FileSystem | None = None,
-    destination_filesystem: FileSystem | None = None,
+    source_filesystem: SupportedFileSystem | None = None,
+    destination_filesystem: SupportedFileSystem | None = None,
     *,
     chunk_size: int = 1024 * 1024,
     use_threads: bool = True,
 ) -> None: ...
 
 class FSSpecHandler(FileSystemHandler):  # type: ignore[misc]
-    fs: FileSystem
-    def __init__(self, fs: FileSystem) -> None: ...
+    fs: SupportedFileSystem
+    def __init__(self, fs: SupportedFileSystem) -> None: ...
 
 __all__ = [
     # _fs
diff --git a/pyarrow-stubs/orc.pyi b/pyarrow-stubs/orc.pyi
index 1b2d277..c0104f1 100644
--- a/pyarrow-stubs/orc.pyi
+++ b/pyarrow-stubs/orc.pyi
@@ -1,7 +1,7 @@
 from typing import IO, Literal, Self
 
 from . import _orc
-from ._fs import FileSystem
+from ._fs import SupportedFileSystem
 from .lib import KeyValueMetadata, NativeFile, RecordBatch, Schema, Table
 
 class ORCFile:
@@ -71,7 +71,7 @@ class ORCWriter:
 def read_table(
     source: str | NativeFile | IO,
     columns: list[str] | None = None,
-    filesystem: FileSystem | None = None,
+    filesystem: SupportedFileSystem | None = None,
 ) -> Table: ...
 def write_table(
     table: Table,
diff --git a/pyarrow-stubs/parquet/core.pyi b/pyarrow-stubs/parquet/core.pyi
index 1611c82..2285bc5 100644
--- a/pyarrow-stubs/parquet/core.pyi
+++ b/pyarrow-stubs/parquet/core.pyi
@@ -3,7 +3,7 @@ from typing import IO, Callable, Iterator, Literal, Self, Sequence
 
 from pyarrow import _parquet
 from pyarrow._compute import Expression
-from pyarrow._fs import FileSystem
+from pyarrow._fs import FileSystem, SupportedFileSystem
 from pyarrow._parquet import (
     ColumnChunkMetaData,
     ColumnSchema,
@@ -70,7 +70,7 @@ class ParquetFile:
         decryption_properties: FileDecryptionProperties | None = None,
         thrift_string_size_limit: int | None = None,
         thrift_container_size_limit: int | None = None,
-        filesystem: FileSystem | None = None,
+        filesystem: SupportedFileSystem | None = None,
         page_checksum_verification: bool = False,
     ): ...
     def __enter__(self) -> Self: ...
@@ -129,7 +129,7 @@ class ParquetWriter:
         self,
         where: str | Path | IO,
         schema: Schema,
-        filesystem: FileSystem | None = None,
+        filesystem: SupportedFileSystem | None = None,
         flavor: str | None = None,
         version: Literal["1.0", "2.4", "2.6"] = ...,
         use_dictionary: bool = True,
@@ -166,7 +166,7 @@ class ParquetDataset:
     def __init__(
         self,
         path_or_paths: str | list[str],
-        filesystem: FileSystem | None = None,
+        filesystem: SupportedFileSystem | None = None,
         schema: Schema | None = None,
         *,
         filters: Expression | FilterTuple | list[FilterTuple] | None = None,
@@ -213,7 +213,7 @@ def read_table(
     memory_map: bool = False,
     buffer_size: int = 0,
     partitioning: str | list[str] | Partitioning = "hive",
-    filesystem: FileSystem | None = None,
+    filesystem: SupportedFileSystem | None = None,
     filters: Expression | FilterTuple | list[FilterTuple] | None = None,
     use_legacy_dataset: bool | None = None,
     ignore_prefixes: list[str] | None = None,
@@ -240,7 +240,7 @@ def write_table(
     allow_truncated_timestamps: bool = False,
     data_page_size: int | None = None,
     flavor: str | None = None,
-    filesystem: FileSystem | None = None,
+    filesystem: SupportedFileSystem | None = None,
     compression_level: int | dict | None = None,
     use_byte_stream_split: bool = False,
     column_encoding: str | dict | None = None,
@@ -260,7 +260,7 @@ def write_to_dataset(
     table: Table,
     root_path: str | Path,
     partition_cols: list[str] | None = None,
-    filesystem: FileSystem | None = None,
+    filesystem: SupportedFileSystem | None = None,
     use_legacy_dataset: bool | None = None,
     schema: Schema | None = None,
     partitioning: Partitioning | list[str] | None = None,
@@ -275,18 +275,18 @@ def write_metadata(
     schema: Schema,
     where: str | NativeFile,
     metadata_collector: list[FileMetaData] | None = None,
-    filesystem: FileSystem | None = None,
+    filesystem: SupportedFileSystem | None = None,
     **kwargs,
 ) -> None: ...
 def read_metadata(
     where: str | Path | IO,
     memory_map: bool = False,
     decryption_properties: FileDecryptionProperties | None = None,
-    filesystem: FileSystem | None = None,
+    filesystem: SupportedFileSystem | None = None,
 ) -> FileMetaData: ...
 def read_schema(
     where: str | Path | IO,
     memory_map: bool = False,
     decryption_properties: FileDecryptionProperties | None = None,
-    filesystem: FileSystem | None = None,
+    filesystem: SupportedFileSystem | None = None,
 ) -> FileMetaData: ...