diff --git a/src/python/nimbusml.pyproj b/src/python/nimbusml.pyproj
index acd5a6c6..49492026 100644
--- a/src/python/nimbusml.pyproj
+++ b/src/python/nimbusml.pyproj
@@ -583,6 +583,7 @@
     <Compile Include="nimbusml\tests\ensemble\__init__.py" />
     <Compile Include="nimbusml\tests\feature_extraction\text\test_sentiment.py" />
     <Compile Include="nimbusml\tests\idv\__init__.py" />
+    <Compile Include="nimbusml\tests\pipeline\test_pipeline_subclassing.py" />
     <Compile Include="nimbusml\tests\preprocessing\normalization\test_meanvariancescaler.py" />
     <Compile Include="nimbusml\tests\timeseries\test_iidchangepointdetector.py" />
     <Compile Include="nimbusml\tests\timeseries\test_ssachangepointdetector.py" />
diff --git a/src/python/nimbusml/pipeline.py b/src/python/nimbusml/pipeline.py
index 7237ef7a..74435d87 100644
--- a/src/python/nimbusml/pipeline.py
+++ b/src/python/nimbusml/pipeline.py
@@ -1839,7 +1839,7 @@ def predict_proba(self, X, verbose=0, **params):
             last_node = self.last_node
             last_node._check_implements_method('predict_proba')
 
-        scores = self.predict(X, verbose, **params)
+        scores, _ = self._predict(X, verbose=verbose, **params)
 
         # REVIEW: Consider adding an entry point that extracts the
         # probability column instead.
@@ -1883,7 +1883,7 @@ def decision_function(self, X, verbose=0, **params):
             last_node = self.last_node
             last_node._check_implements_method('decision_function')
 
-        scores = self.predict(X, verbose, **params)
+        scores, _ = self._predict(X, verbose=verbose, **params)
 
         # REVIEW: Consider adding an entry point that extracts the score
         #  column instead.
diff --git a/src/python/nimbusml/tests/pipeline/test_pipeline_subclassing.py b/src/python/nimbusml/tests/pipeline/test_pipeline_subclassing.py
new file mode 100644
index 00000000..56d9ef42
--- /dev/null
+++ b/src/python/nimbusml/tests/pipeline/test_pipeline_subclassing.py
@@ -0,0 +1,73 @@
+# --------------------------------------------------------------------------------------------
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# Licensed under the MIT License.
+# --------------------------------------------------------------------------------------------
+import os
+import unittest
+
+import numpy as np
+import pandas as pd
+from nimbusml import Pipeline
+from nimbusml.linear_model import LogisticRegressionBinaryClassifier
+
+
+def generate_dataset_1():
+    X = pd.DataFrame({'x1': [2, 3, 2, 2, 8, 9, 10, 8],
+                      'x2': [1, 2, 3, 1, 7, 10, 9, 8]})
+    y = pd.DataFrame({'y': [1, 1, 1, 1, 0, 0, 0, 0]})
+    return X, y
+
+
+class CustomPipeline(Pipeline):
+    # Override the predict method
+    def predict(self, X, *args, **kwargs):
+        return kwargs.get('test_return_value')
+
+
+class TestPipelineSubclassing(unittest.TestCase):
+
+    def test_pipeline_subclass_can_override_predict(self):
+        X, y = generate_dataset_1()
+
+        pipeline = Pipeline([LogisticRegressionBinaryClassifier()])
+        pipeline.fit(X, y)
+        result = pipeline.predict(X)['PredictedLabel']
+
+        self.assertTrue(np.array_equal(result.values, y['y'].values))
+
+        pipeline = CustomPipeline([LogisticRegressionBinaryClassifier()])
+        pipeline.fit(X, y)
+
+        self.assertEqual(pipeline.predict(X, test_return_value=3), 3)
+
+
+    def test_pipeline_subclass_correctly_supports_predict_proba(self):
+        X, y = generate_dataset_1()
+
+        pipeline = Pipeline([LogisticRegressionBinaryClassifier()])
+        pipeline.fit(X, y)
+        orig_result = pipeline.predict_proba(X)
+
+        pipeline = CustomPipeline([LogisticRegressionBinaryClassifier()])
+        pipeline.fit(X, y)
+        new_result = pipeline.predict_proba(X)
+
+        self.assertTrue(np.array_equal(orig_result, new_result))
+
+
+    def test_pipeline_subclass_correctly_supports_decision_function(self):
+        X, y = generate_dataset_1()
+
+        pipeline = Pipeline([LogisticRegressionBinaryClassifier()])
+        pipeline.fit(X, y)
+        orig_result = pipeline.decision_function(X)
+
+        pipeline = CustomPipeline([LogisticRegressionBinaryClassifier()])
+        pipeline.fit(X, y)
+        new_result = pipeline.decision_function(X)
+
+        self.assertTrue(np.array_equal(orig_result, new_result))
+
+
+if __name__ == '__main__':
+    unittest.main()