diff --git a/tabulate/__init__.py b/tabulate/__init__.py
index 12a2950..cd12475 100644
--- a/tabulate/__init__.py
+++ b/tabulate/__init__.py
@@ -1483,11 +1483,20 @@ def _normalize_tabular_data(tabular_data, headers, showindex="default"):
         elif hasattr(tabular_data, "index"):
             # values is a property, has .index => it's likely a pandas.DataFrame (pandas 0.11.0)
             keys = list(tabular_data)
-            if showindex in ["default", "always", True] and tabular_data.index.name is not None:
-                if isinstance(tabular_data.index.name, list):
-                    keys[:0] = tabular_data.index.name
-                else:
-                    keys[:0] = [tabular_data.index.name]
+            if showindex in ["default", "always", True]:
+                # Pandas prints columns.name in the index header when the
+                # index itself has no name. Keep an explicit index.name.
+                index_header = tabular_data.index.name
+                if index_header is None:
+                    columns = getattr(tabular_data, "columns", None)
+                    column_axis_name = getattr(columns, "name", None)
+                    if column_axis_name is not None:
+                        index_header = column_axis_name
+                if index_header is not None:
+                    if isinstance(index_header, list):
+                        keys[:0] = index_header
+                    else:
+                        keys[:0] = [index_header]
             vals = tabular_data.values  # values matrix doesn't need to be transposed
             # for DataFrames add an index per default
             index = list(tabular_data.index)
diff --git a/test/test_input.py b/test/test_input.py
index 3cc3237..7c551ab 100644
--- a/test/test_input.py
+++ b/test/test_input.py
@@ -292,6 +292,33 @@ def test_pandas_keys():
         skip("test_pandas_keys is skipped")
 
 
+def test_pandas_columns_name_shown_when_index_unnamed():
+    "DataFrame columns.name fills the index header when index.name is unset."
+    try:
+        import pandas
+
+        df = pandas.DataFrame([[1, 2], [3, 4]], columns=["a", "b"])
+        df.columns.name = "some_name"
+        expected = """\
+|   some_name |   a |   b |
+|------------:|----:|----:|
+|           0 |   1 |   2 |
+|           1 |   3 |   4 |"""
+        result = tabulate(df, tablefmt="pipe", headers="keys", showindex=True)
+        assert_equal(expected, result)
+
+        df.index.name = "row_id"
+        named_index = tabulate(df, tablefmt="pipe", headers="keys", showindex=True)
+        assert_in("row_id", named_index.splitlines()[0])
+        assert "some_name" not in named_index.splitlines()[0]
+
+        hidden = tabulate(df, tablefmt="pipe", headers="keys", showindex=False)
+        assert "some_name" not in hidden
+        assert "row_id" not in hidden
+    except ImportError:
+        skip("test_pandas_columns_name_shown_when_index_unnamed is skipped")
+
+
 def test_sqlite3():
     "Input: an sqlite3 cursor"
     try: