diff --git a/python-package/lightgbm/basic.py b/python-package/lightgbm/basic.py
index ba704a06c360..4c32add1d3e6 100644
--- a/python-package/lightgbm/basic.py
+++ b/python-package/lightgbm/basic.py
@@ -23,6 +23,12 @@
 _DatasetHandle = ctypes.c_void_p
 _LGBM_EvalFunctionResultType = Tuple[str, float, bool]
 _LGBM_BoosterEvalMethodResultType = Tuple[str, str, float, bool]
+_LGBM_LabelType = Union[
+    list,
+    np.ndarray,
+    pd_Series,
+    pd_DataFrame
+]
 
 ZERO_THRESHOLD = 1e-35
 
@@ -605,15 +611,6 @@ def _data_from_pandas(data, feature_name, categorical_feature, pandas_categorica
     return data, feature_name, categorical_feature, pandas_categorical
 
 
-def _label_from_pandas(label):
-    if isinstance(label, pd_DataFrame):
-        if len(label.columns) > 1:
-            raise ValueError('DataFrame for label cannot have multiple columns')
-        _check_for_bad_pandas_dtypes(label.dtypes)
-        label = np.ravel(label.values.astype(np.float32, copy=False))
-    return label
-
-
 def _dump_pandas_categorical(pandas_categorical, file_name=None):
     categorical_json = json.dumps(pandas_categorical, default=json_default_with_numpy)
     pandas_str = f'\npandas_categorical:{categorical_json}\n'
@@ -1200,7 +1197,7 @@ class Dataset:
     def __init__(
         self,
         data,
-        label=None,
+        label: Optional[_LGBM_LabelType] = None,
         reference: Optional["Dataset"] = None,
         weight=None,
         group=None,
@@ -1505,7 +1502,7 @@ def _set_init_score_by_predictor(self, predictor, data, used_indices=None):
     def _lazy_init(
         self,
         data,
-        label=None,
+        label: Optional[_LGBM_LabelType] = None,
         reference: Optional["Dataset"] = None,
         weight=None,
         group=None,
@@ -1525,7 +1522,6 @@ def _lazy_init(
                                                                                              feature_name,
                                                                                              categorical_feature,
                                                                                              self.pandas_categorical)
-        label = _label_from_pandas(label)
 
         # process for args
         params = {} if params is None else params
@@ -1936,7 +1932,7 @@ def construct(self) -> "Dataset":
     def create_valid(
         self,
         data,
-        label=None,
+        label: Optional[_LGBM_LabelType] = None,
         weight=None,
         group=None,
         init_score=None,
@@ -2276,7 +2272,7 @@ def set_feature_name(self, feature_name: List[str]) -> "Dataset":
                 ctypes.c_int(len(feature_name))))
         return self
 
-    def set_label(self, label) -> "Dataset":
+    def set_label(self, label: Optional[_LGBM_LabelType]) -> "Dataset":
         """Set label of Dataset.
 
         Parameters
@@ -2291,8 +2287,14 @@ def set_label(self, label) -> "Dataset":
         """
         self.label = label
         if self.handle is not None:
-            label = list_to_1d_numpy(_label_from_pandas(label), name='label')
-            self.set_field('label', label)
+            if isinstance(label, pd_DataFrame):
+                if len(label.columns) > 1:
+                    raise ValueError('DataFrame for label cannot have multiple columns')
+                _check_for_bad_pandas_dtypes(label.dtypes)
+                label_array = np.ravel(label.values.astype(np.float32, copy=False))
+            else:
+                label_array = list_to_1d_numpy(label, name='label')
+            self.set_field('label', label_array)
             self.label = self.get_field('label')  # original values can be modified at cpp side
         return self