diff --git a/openml/datasets/functions.py b/openml/datasets/functions.py
index 12ad42784..fcd29515d 100644
--- a/openml/datasets/functions.py
+++ b/openml/datasets/functions.py
@@ -742,6 +742,8 @@ def create_dataset(  # noqa: C901, PLR0912, PLR0915
 
     data_format: Literal["arff", "sparse_arff"]
     if isinstance(data, (list, np.ndarray)):
+        if len(data) == 0:
+            raise ValueError("'data' must contain at least one row, but it is empty.")
         if isinstance(data[0], (list, np.ndarray)):
             data_format = "arff"
         elif isinstance(data[0], dict):
diff --git a/tests/test_datasets/test_dataset_functions.py b/tests/test_datasets/test_dataset_functions.py
index fae4a1c4e..f5053ea23 100644
--- a/tests/test_datasets/test_dataset_functions.py
+++ b/tests/test_datasets/test_dataset_functions.py
@@ -877,6 +877,10 @@ def test_create_invalid_dataset(self):
         param["data"] = data[0]
         self.assertRaises(ValueError, create_dataset, **param)  # noqa: PT027
 
+        param["data"] = np.empty((0, 2))
+        with pytest.raises(ValueError, match="at least one row"):
+            create_dataset(**param)
+
     @pytest.mark.test_server
     def test_get_online_dataset_arff(self):
         dataset_id = 128  # iris -- one of the few datasets without parquet file