diff --git a/python/pyspark/ml/classification.py b/python/pyspark/ml/classification.py
index 1392bc763dd94..5523b6cde7157 100644
--- a/python/pyspark/ml/classification.py
+++ b/python/pyspark/ml/classification.py
@@ -30,7 +30,7 @@
 from pyspark.ml.wrapper import JavaParams, \
     JavaPredictor, _JavaPredictorParams, JavaPredictionModel, JavaWrapper
 from pyspark.ml.common import inherit_doc
-from pyspark.ml.linalg import Vectors
+from pyspark.ml.linalg import Vectors, VectorUDT
 from pyspark.sql import DataFrame
 from pyspark.sql.functions import udf, when
 from pyspark.sql.types import ArrayType, DoubleType
@@ -2724,7 +2724,7 @@ def func(predictions):
                     predArray.append(x)
                 return Vectors.dense(predArray)
 
-            rawPredictionUDF = udf(func)
+            rawPredictionUDF = udf(func, VectorUDT())
             aggregatedDataset = aggregatedDataset.withColumn(
                 self.getRawPredictionCol(), rawPredictionUDF(aggregatedDataset[accColName]))
 
diff --git a/python/pyspark/ml/tests/test_algorithms.py b/python/pyspark/ml/tests/test_algorithms.py
index 2faf2d98f0271..90fe59f3ae800 100644
--- a/python/pyspark/ml/tests/test_algorithms.py
+++ b/python/pyspark/ml/tests/test_algorithms.py
@@ -25,7 +25,7 @@
     MultilayerPerceptronClassifier, OneVsRest
 from pyspark.ml.clustering import DistributedLDAModel, KMeans, LocalLDAModel, LDA, LDAModel
 from pyspark.ml.fpm import FPGrowth
-from pyspark.ml.linalg import Matrices, Vectors
+from pyspark.ml.linalg import Matrices, Vectors, DenseVector
 from pyspark.ml.recommendation import ALS
 from pyspark.ml.regression import GeneralizedLinearRegression, LinearRegression
 from pyspark.sql import Row
@@ -116,7 +116,19 @@ def test_output_columns(self):
         output = model.transform(df)
         self.assertEqual(output.columns, ["label", "features", "rawPrediction", "prediction"])
 
-    def test_parallelism_doesnt_change_output(self):
+    def test_raw_prediction_column_is_of_vector_type(self):
+        # SPARK-35142: `OneVsRestModel` outputs raw prediction as a string column
+        df = self.spark.createDataFrame([(0.0, Vectors.dense(1.0, 0.8)),
+                                         (1.0, Vectors.sparse(2, [], [])),
+                                         (2.0, Vectors.dense(0.5, 0.5))],
+                                        ["label", "features"])
+        lr = LogisticRegression(maxIter=5, regParam=0.01)
+        ovr = OneVsRest(classifier=lr, parallelism=1)
+        model = ovr.fit(df)
+        row = model.transform(df).head()
+        self.assertIsInstance(row["rawPrediction"], DenseVector)
+
+    def test_parallelism_does_not_change_output(self):
         df = self.spark.createDataFrame([(0.0, Vectors.dense(1.0, 0.8)),
                                          (1.0, Vectors.sparse(2, [], [])),
                                          (2.0, Vectors.dense(0.5, 0.5))],