diff --git a/autosklearn/automl.py b/autosklearn/automl.py
index 09944bec66..f751e87a0a 100644
--- a/autosklearn/automl.py
+++ b/autosklearn/automl.py
@@ -943,7 +943,7 @@ def predict(self, X, batch_size=None, n_jobs=1):
             return predicted_classes
 
     def predict_proba(self, X, batch_size=None, n_jobs=1):
-        return self._automl.predict(X, batch_size=batch_size, n_jobs=n_jobs)
+        return super().predict(X, batch_size=batch_size, n_jobs=n_jobs)
 
 
 class AutoMLRegressor(BaseAutoML):
diff --git a/autosklearn/estimators.py b/autosklearn/estimators.py
index 1cd600ee2e..96f0a3a5b5 100644
--- a/autosklearn/estimators.py
+++ b/autosklearn/estimators.py
@@ -308,6 +308,10 @@ def refit(self, X, y):
     def predict(self, X, batch_size=None, n_jobs=1):
         return self._automl.predict(X, batch_size=batch_size, n_jobs=n_jobs)
 
+    def predict_proba(self, X, batch_size=None, n_jobs=1):
+        return self._automl.predict_proba(
+             X, batch_size=batch_size, n_jobs=n_jobs)
+
     def score(self, X, y):
         return self._automl.score(X, y)
 
@@ -428,7 +432,7 @@ def predict_proba(self, X, batch_size=None, n_jobs=1):
             The predicted class probabilities.
 
         """
-        return self._automl.predict_proba(
+        return super().predict_proba(
             X, batch_size=batch_size, n_jobs=n_jobs)
 
 
diff --git a/test/test_automl/test_estimators.py b/test/test_automl/test_estimators.py
index 0ba9d68233..2e23f00dd1 100644
--- a/test/test_automl/test_estimators.py
+++ b/test/test_automl/test_estimators.py
@@ -103,7 +103,6 @@ def test_fit_pSMAC(self):
                                        initial_configurations_via_metalearning=0,
                                        ensemble_size=0)
         automl.fit(X_train, Y_train)
-
         # Create a 'dummy model' for the first run, which has an accuracy of
         # more than 99%; it should be in the final ensemble if the ensemble
         # building of the second AutoSklearn classifier works correct
@@ -303,6 +302,8 @@ def test_multilabel(self):
         self.assertEqual(predictions.shape, (50, 3))
         score = f1_macro(Y_test, predictions)
         self.assertGreaterEqual(score, 0.9)
+        probs = automl.predict_proba(X_train)
+        self.assertAlmostEqual(np.mean(probs), 0.33333333333333331)
 
     def test_binary(self):
         output = os.path.join(self.test_dir, '..', '.tmp_binary_fit')
@@ -376,4 +377,4 @@ def test_conversion_of_list_to_np(self, fit_ensemble, refit, fit):
         self.assertIsInstance(refit.call_args[0][1], np.ndarray)
         automl.fit_ensemble(y)
         self.assertEqual(fit_ensemble.call_count, 1)
-        self.assertIsInstance(fit_ensemble.call_args[0][0], np.ndarray)
\ No newline at end of file
+        self.assertIsInstance(fit_ensemble.call_args[0][0], np.ndarray)