diff --git a/doc/source/whatsnew/v1.0.0.rst b/doc/source/whatsnew/v1.0.0.rst
index 16d23d675a8bb1..f8c4f9f3dc4102 100644
--- a/doc/source/whatsnew/v1.0.0.rst
+++ b/doc/source/whatsnew/v1.0.0.rst
@@ -191,7 +191,7 @@ Datetimelike
 - Bug in :class:`Series` and :class:`DataFrame` with integer dtype failing to raise ``TypeError`` when adding or subtracting a ``np.datetime64`` object (:issue:`28080`)
 - Bug in :class:`Week` with ``weekday`` incorrectly raising ``AttributeError`` instead of ``TypeError`` when adding or subtracting an invalid type (:issue:`28530`)
 - Bug in :class:`DataFrame` arithmetic operations when operating with a :class:`Series` with dtype `'timedelta64[ns]'` (:issue:`28049`)
--
+- Bug in :func:`pandas.core.groupby.generic.SeriesGroupBy.apply` raising ``ValueError`` when a column in the original DataFrame is a datetime and the column labels are not standard integers (:issue:`28247`)
 
 Timedelta
 ^^^^^^^^^
diff --git a/pandas/core/groupby/generic.py b/pandas/core/groupby/generic.py
index b5aec189700ce8..e556708dc92836 100644
--- a/pandas/core/groupby/generic.py
+++ b/pandas/core/groupby/generic.py
@@ -1913,7 +1913,9 @@ def _recast_datetimelike_result(result: DataFrame) -> DataFrame:
     result = result.copy()
 
     obj_cols = [
-        idx for idx in range(len(result.columns)) if is_object_dtype(result.dtypes[idx])
+        idx
+        for idx in range(len(result.columns))
+        if is_object_dtype(result.dtypes.iloc[idx])
     ]
 
     # See GH#26285
diff --git a/pandas/tests/groupby/test_apply.py b/pandas/tests/groupby/test_apply.py
index 76588549532b1e..4d0063b773bc5c 100644
--- a/pandas/tests/groupby/test_apply.py
+++ b/pandas/tests/groupby/test_apply.py
@@ -657,3 +657,22 @@ def test_apply_with_mixed_types():
 
     result = g.apply(lambda x: x / x.sum())
     tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "group_column_dtlike",
+    [datetime.today(), datetime.today().date(), datetime.today().time()],
+)
+def test_apply_datetime_issue(group_column_dtlike):
+    # GH-28247
+    # groupby-apply throws an error if one of the columns in the DataFrame
+    #   is a datetime object and the column labels are different from
+    #   standard int values in range(len(num_columns))
+
+    df = pd.DataFrame({"a": ["foo"], "b": [group_column_dtlike]})
+    result = df.groupby("a").apply(lambda x: pd.Series(["spam"], index=[42]))
+
+    expected = pd.DataFrame(
+        ["spam"], Index(["foo"], dtype="object", name="a"), columns=[42]
+    )
+    tm.assert_frame_equal(result, expected)