diff --git a/tensorflow/python/keras/engine/training.py b/tensorflow/python/keras/engine/training.py
index a0ebec4f95e..5355920ced5 100644
--- a/tensorflow/python/keras/engine/training.py
+++ b/tensorflow/python/keras/engine/training.py
@@ -1592,6 +1592,7 @@ class Model(base_layer.Layer, version_utils.ModelVersionSelector):
       predict_function = self.make_predict_function()
       self._predict_counter.assign(0)
       callbacks.on_predict_begin()
+      batch_outputs = None
       for _, iterator in data_handler.enumerate_epochs():  # Single epoch.
         with data_handler.catch_stop_iteration():
           for step in data_handler.steps():
@@ -1610,6 +1611,8 @@ class Model(base_layer.Layer, version_utils.ModelVersionSelector):
                   outputs, batch_outputs)
             end_step = step + data_handler.step_increment
             callbacks.on_predict_batch_end(end_step, {'outputs': batch_outputs})
+      if batch_outputs is None:
+        raise ValueError('Expect x to be a non-empty array or dataset.')
       callbacks.on_predict_end()
     all_outputs = nest.map_structure_up_to(batch_outputs, concat, outputs)
     return tf_utils.to_numpy_or_python_type(all_outputs)
diff --git a/tensorflow/python/keras/engine/training_test.py b/tensorflow/python/keras/engine/training_test.py
index 5cf15926bfb..ad904ce9aa7 100644
--- a/tensorflow/python/keras/engine/training_test.py
+++ b/tensorflow/python/keras/engine/training_test.py
@@ -1647,6 +1647,19 @@ class TestExceptionsAndWarnings(keras_parameterized.TestCase):
     ):
       training_module.Model([inputs], output)
 
+  @keras_parameterized.run_all_keras_modes(always_skip_v1=True)
+  def test_predict_error_with_empty_x(self):
+    inputs = layers_module.Input(shape=(2,))
+    outputs = layers_module.Dense(4)(inputs)
+    model = training_module.Model(inputs=inputs, outputs=outputs)
+    model.compile(loss='mse')
+
+    with self.assertRaisesRegexp(
+        ValueError,
+        'Expect x to be a non-empty array or dataset.'
+    ):
+      model.predict(np.array([]))
+
 
 class LossWeightingTest(keras_parameterized.TestCase):