BUG: added missing fill_na parameter to DataFrame.unstack() with list of levels

doc/source/whatsnew/v0.25.3.rst

@@ -10,6 +10,7 @@ including other versions of pandas.
 
 Bug fixes
 ~~~~~~~~~
+- Bug in missing fill_na parameter to DataFrame.unstack() with list of levels (:issue:`30740`)
 
 Groupby/resample/rolling
 ^^^^^^^^^^^^^^^^^^^^^^^^

pandas/core/reshape/reshape.py

@@ -358,7 +358,7 @@ def _unstack_multiple(data, clocs, fill_value=None):
             result = data
             for i in range(len(clocs)):
                 val = clocs[i]
-                result = result.unstack(val)
+                result = result.unstack(val, fill_value)
                 clocs = [v if i > v else v - 1 for v in clocs]
 
             return result

pandas/tests/frame/test_reshape.py

@@ -1128,3 +1128,22 @@ def test_stack_timezone_aware_values():
         ),
     )
     tm.assert_series_equal(result, expected)
+
+
+class TestMultiIndexReshape:
+    def test_unstacking_multi_index_df(self):
+        # BUG: gh-30740
+        df = pd.DataFrame(
+            {
+                "name": ["Alice", "Bob"],
+                "score": [9.5, 8],
+                "employed": [False, True],
+                "kids": [0, 0],
+                "gender": ["female", "male"],
+            }
+        )
+        df = df.set_index(["name", "employed", "kids", "gender"])
+        df = df.unstack(["gender"], fill_value=0)
+        expected = df.unstack("employed", fill_value=0).unstack("kids", fill_value=0)
+        result = df.unstack(["employed", "kids"], fill_value=0)
+        tm.assert_frame_equal(result, expected)