diff --git a/mmcls/datasets/base_dataset.py b/mmcls/datasets/base_dataset.py
index 3d4212d2e..9fe3bf03f 100644
--- a/mmcls/datasets/base_dataset.py
+++ b/mmcls/datasets/base_dataset.py
@@ -1,6 +1,7 @@
 import copy
 from abc import ABCMeta, abstractmethod
 
+import mmcv
 import numpy as np
 from torch.utils.data import Dataset
 
@@ -23,7 +24,12 @@ class BaseDataset(Dataset, metaclass=ABCMeta):
 
     CLASSES = None
 
-    def __init__(self, data_prefix, pipeline, ann_file=None, test_mode=False):
+    def __init__(self,
+                 data_prefix,
+                 pipeline,
+                 classes=None,
+                 ann_file=None,
+                 test_mode=False):
         super(BaseDataset, self).__init__()
 
         self.ann_file = ann_file
@@ -31,6 +37,7 @@ class BaseDataset(Dataset, metaclass=ABCMeta):
         self.test_mode = test_mode
         self.pipeline = Compose(pipeline)
         self.data_infos = self.load_annotations()
+        self.CLASSES = self.get_classes(classes)
 
     @abstractmethod
     def load_annotations(self):
@@ -78,6 +85,32 @@ class BaseDataset(Dataset, metaclass=ABCMeta):
     def __getitem__(self, idx):
         return self.prepare_data(idx)
 
+    @classmethod
+    def get_classes(cls, classes=None):
+        """Get class names of current dataset.
+        Args:
+            classes (Sequence[str] | str | None): If classes is None, use
+                default CLASSES defined by builtin dataset. If classes is a
+                string, take it as a file name. The file contains the name of
+                classes where each line contains one class name. If classes is
+                a tuple or list, override the CLASSES defined by the dataset.
+
+        Returns:
+            tuple[str] or list[str]: Names of categories of the dataset.
+        """
+        if classes is None:
+            return cls.CLASSES
+
+        if isinstance(classes, str):
+            # take it as a file path
+            class_names = mmcv.list_from_file(classes)
+        elif isinstance(classes, (tuple, list)):
+            class_names = classes
+        else:
+            raise ValueError(f'Unsupported type {type(classes)} of classes.')
+
+        return class_names
+
     def evaluate(self,
                  results,
                  metric='accuracy',
diff --git a/tests/test_dataset.py b/tests/test_dataset.py
index d4d930d23..17fed0eb3 100644
--- a/tests/test_dataset.py
+++ b/tests/test_dataset.py
@@ -21,12 +21,44 @@ def test_datasets_override_default(dataset_name):
     dataset_class = DATASETS.get(dataset_name)
     dataset_class.load_annotations = MagicMock()
 
+    original_classes = dataset_class.CLASSES
+
+    # Test setting classes as a tuple
+    dataset = dataset_class(
+        data_prefix='', pipeline=[], classes=('bus', 'car'), test_mode=True)
+    assert dataset.CLASSES != original_classes
+    assert dataset.CLASSES == ('bus', 'car')
+
+    # Test setting classes as a list
+    dataset = dataset_class(
+        data_prefix='', pipeline=[], classes=['bus', 'car'], test_mode=True)
+    assert dataset.CLASSES != original_classes
+    assert dataset.CLASSES == ['bus', 'car']
+
+    # Test setting classes through a file
+    tmp_file = tempfile.NamedTemporaryFile()
+    with open(tmp_file.name, 'w') as f:
+        f.write('bus\ncar\n')
+    dataset = dataset_class(
+        data_prefix='', pipeline=[], classes=tmp_file.name, test_mode=True)
+    tmp_file.close()
+
+    assert dataset.CLASSES != original_classes
+    assert dataset.CLASSES == ['bus', 'car']
+
+    # Test overriding not a subset
+    dataset = dataset_class(
+        data_prefix='', pipeline=[], classes=['foo'], test_mode=True)
+    assert dataset.CLASSES != original_classes
+    assert dataset.CLASSES == ['foo']
+
     # Test default behavior
     dataset = dataset_class(data_prefix='', pipeline=[])
 
     assert dataset.data_prefix == ''
     assert not dataset.test_mode
     assert dataset.ann_file is None
+    assert dataset.CLASSES == original_classes
 
 
 @patch.multiple(BaseDataset, __abstractmethods__=set())