diff --git a/doc/source/whatsnew/v1.0.0.rst b/doc/source/whatsnew/v1.0.0.rst
index 4f0ca97310d85..2c2f0ad4c006f 100755
--- a/doc/source/whatsnew/v1.0.0.rst
+++ b/doc/source/whatsnew/v1.0.0.rst
@@ -1273,6 +1273,7 @@ Other
 - Bug in :meth:`DataFrame.append` that raised ``IndexError`` when appending with empty list (:issue:`28769`)
 - Fix :class:`AbstractHolidayCalendar` to return correct results for
   years after 2030 (now goes up to 2200) (:issue:`27790`)
+- Bug in :meth:`DataFrame.apply` returning wrong result in some cases when dtype was involved in passed function (:issue:`28773`)
 - Fixed :class:`~arrays.IntegerArray` returning ``inf`` rather than ``NaN`` for operations dividing by ``0`` (:issue:`27398`)
 - Fixed ``pow`` operations for :class:`~arrays.IntegerArray` when the other value is ``0`` or ``1`` (:issue:`29997`)
 - Bug in :meth:`Series.count` raises if use_inf_as_na is enabled (:issue:`29478`)
diff --git a/pandas/core/apply.py b/pandas/core/apply.py
index a013434491589..90c1f20b7da8a 100644
--- a/pandas/core/apply.py
+++ b/pandas/core/apply.py
@@ -14,7 +14,10 @@
     is_list_like,
     is_sequence,
 )
-from pandas.core.dtypes.generic import ABCSeries
+from pandas.core.dtypes.generic import (
+    ABCSeries,
+    ABCMultiIndex,
+)
 
 from pandas.core.construction import create_series_with_explicit_dtype
 
@@ -271,16 +274,22 @@ def apply_standard(self):
 
         # we cannot reduce using non-numpy dtypes,
         # as demonstrated in gh-12244
-        if (
+        can_reduce = (
             self.result_type in ["reduce", None]
             and not self.dtypes.apply(is_extension_array_dtype).any()
+            # Disallow complex_internals since libreduction shortcut
+            #  cannot handle MultiIndex
+            and not isinstance(self.agg_axis, ABCMultiIndex)
             # Disallow dtypes where setting _index_data will break
             #  ExtensionArray values, see GH#31182
             and not self.dtypes.apply(lambda x: x.kind in ["m", "M"]).any()
             # Disallow complex_internals since libreduction shortcut raises a TypeError
             and not self.agg_axis._has_complex_internals
-        ):
+        )
+
+        column_by_column = (self.axis != 0 and self.axis != "index") or self.obj._is_homogeneous_type
 
+        if can_reduce and column_by_column:
             values = self.values
             index = self.obj._get_axis(self.axis)
             labels = self.agg_axis
@@ -309,9 +318,17 @@ def apply_standard(self):
             else:
                 return self.obj._constructor_sliced(result, index=labels)
 
+
         # compute the result using the series generator
         results, res_index = self.apply_series_generator()
 
+        if can_reduce and not column_by_column:
+            results = list(results.values())
+            results = np.array(results)
+            return self.obj._constructor_sliced(
+                results, index=res_index
+            )
+
         # wrap results
         return self.wrap_results(results, res_index)
 
diff --git a/pandas/tests/frame/test_apply.py b/pandas/tests/frame/test_apply.py
index e328523253144..34267c413920a 100644
--- a/pandas/tests/frame/test_apply.py
+++ b/pandas/tests/frame/test_apply.py
@@ -704,6 +704,13 @@ def test_apply_dup_names_multi_agg(self):
 
         tm.assert_frame_equal(result, expected)
 
+    def test_apply_get_dtype(self):
+        # GH 28773
+        df = DataFrame({"col_1": [1, 2, 3], "col_2": ["hi", "there", "friend"]})
+        result = df.apply(lambda x: x.dtype)
+        expected = Series(data=["int64", "object"], index=["col_1", "col_2"])
+        tm.assert_series_equal(result, expected)
+
     def test_apply_nested_result_axis_1(self):
         # GH 13820
         def apply_list(row):