diff --git a/doc/source/whatsnew/v3.0.0.rst b/doc/source/whatsnew/v3.0.0.rst
index a95f0485abd5f..6b412f5427811 100644
--- a/doc/source/whatsnew/v3.0.0.rst
+++ b/doc/source/whatsnew/v3.0.0.rst
@@ -324,6 +324,7 @@ MultiIndex
 
 I/O
 ^^^
+- Bug in :meth:`DataFrame.to_excel` when writing empty :class:`DataFrame` with :class:`MultiIndex` on both axes (:issue:`57696`)
 -
 -
 
diff --git a/pandas/io/formats/excel.py b/pandas/io/formats/excel.py
index 504b5aa560670..ee7739df49389 100644
--- a/pandas/io/formats/excel.py
+++ b/pandas/io/formats/excel.py
@@ -620,7 +620,7 @@ def _format_header_mi(self) -> Iterable[ExcelCell]:
         lnum = 0
 
         if self.index and isinstance(self.df.index, MultiIndex):
-            coloffset = len(self.df.index[0]) - 1
+            coloffset = self.df.index.nlevels - 1
 
         if self.merge_cells:
             # Format multi-index as a merged cells.
diff --git a/pandas/tests/io/excel/test_writers.py b/pandas/tests/io/excel/test_writers.py
index 6637b4f931102..d3ddc13c1497e 100644
--- a/pandas/tests/io/excel/test_writers.py
+++ b/pandas/tests/io/excel/test_writers.py
@@ -936,6 +936,17 @@ def test_to_excel_empty_multiindex(self, tmp_excel):
             result, expected, check_index_type=False, check_dtype=False
         )
 
+    def test_to_excel_empty_multiindex_both_axes(self, tmp_excel):
+        # GH 57696
+        df = DataFrame(
+            [],
+            index=MultiIndex.from_tuples([], names=[0, 1]),
+            columns=MultiIndex.from_tuples([("A", "B")]),
+        )
+        df.to_excel(tmp_excel)
+        result = pd.read_excel(tmp_excel, header=[0, 1], index_col=[0, 1])
+        tm.assert_frame_equal(result, df)
+
     def test_to_excel_float_format(self, tmp_excel):
         df = DataFrame(
             [[0.123456, 0.234567, 0.567567], [12.32112, 123123.2, 321321.2]],