BUG: Improved error msg

doc/source/whatsnew/v1.1.0.rst

@@ -479,6 +479,7 @@ Indexing
 - Bug in :meth:`DataFrame.copy` _item_cache not invalidated after copy causes post-copy value updates to not be reflected (:issue:`31784`)
 - Bug in `Series.__getitem__` with an integer key and a :class:`MultiIndex` with leading integer level failing to raise ``KeyError`` if the key is not present in the first level (:issue:`33355`)
 - Bug in :meth:`DataFrame.iloc` when slicing a single column-:class:`DataFrame`` with ``ExtensionDtype`` (e.g. ``df.iloc[:, :1]``) returning an invalid result (:issue:`32957`)
+- Calling :meth:`DataFrame.at` on a DataFrame with a MultiIndex raises an exception with a more informative message (:issue:`9259`)
 
 Missing
 ^^^^^^^

pandas/core/indexing.py

@@ -2032,7 +2032,16 @@ def __getitem__(self, key):
                 raise ValueError("Invalid call for scalar access (getting)!")
 
         key = self._convert_key(key)
-        return self.obj._get_value(*key, takeable=self._takeable)
+        try:
+            return self.obj._get_value(*key, takeable=self._takeable)
+        except KeyError as err:
+            if isinstance(self.obj.index, ABCMultiIndex):
+                raise KeyError(
+                    f"Detected KeyError {err}, indexing with {key} "
+                    "failing for MultiIndex"
+                )
+            else:
+                raise
 
     def __setitem__(self, key, value):
         if isinstance(key, tuple):

pandas/tests/indexes/multi/test_get_set.py

@@ -329,3 +329,21 @@ def test_set_levels_with_iterable():
         [expected_sizes, colors], names=["size", "color"]
     )
     tm.assert_index_equal(result, expected)
+
+
+def test_at_indexing_fails_multiindex():
+    # GH9259
+    cols = [("a_col", chr(i + 65)) for i in range(2)]
+    idx = [("a_row", chr(i + 65)) for i in range(2)]
+    df = pd.DataFrame(
+        np.linspace(1, 4, 4).reshape(2, 2),
+        index=pd.MultiIndex.from_tuples(idx),
+        columns=pd.MultiIndex.from_tuples(cols),
+    )
+    s = pd.Series(np.linspace(1, 2, 2), index=pd.MultiIndex.from_tuples(idx))
+
+    with pytest.raises(KeyError, match=r".+? indexing with .+? failing for MultiIndex"):
+        df.at["a_row", "A"]
+
+    with pytest.raises(TypeError, match=r".+? got multiple values for argument .+?"):
+        s.at["a_row", "A"]