diff --git a/python/pyspark/ml/clustering.py b/python/pyspark/ml/clustering.py
index 7267ee2805987..7ccb6db8d1023 100644
--- a/python/pyspark/ml/clustering.py
+++ b/python/pyspark/ml/clustering.py
@@ -51,6 +51,7 @@
 from pyspark.ml.wrapper import JavaEstimator, JavaModel, JavaParams, JavaWrapper
 from pyspark.ml.common import inherit_doc
 from pyspark.ml.stat import MultivariateGaussian
+from pyspark.ml.util import RemoteModelRef
 from pyspark.sql import DataFrame
 from pyspark.ml.linalg import Vector, Matrix
 from pyspark.sql.utils import is_remote
@@ -1581,10 +1582,9 @@ def toLocal(self) -> "LocalLDAModel":
 
         .. warning:: This involves collecting a large :py:func:`topicsMatrix` to the driver.
         """
-        model = LocalLDAModel(self._call_java("toLocal"))
         if is_remote():
-            return model
-
+            return LocalLDAModel(RemoteModelRef(self._call_java("toLocal")))
+        model = LocalLDAModel(self._call_java("toLocal"))
         # SPARK-10931: Temporary fix to be removed once LDAModel defines Params
         model._create_params_from_java()
         model._transfer_params_from_java()
diff --git a/python/pyspark/ml/tests/test_clustering.py b/python/pyspark/ml/tests/test_clustering.py
index 1b8eb73135a96..5ed1315300de9 100644
--- a/python/pyspark/ml/tests/test_clustering.py
+++ b/python/pyspark/ml/tests/test_clustering.py
@@ -378,8 +378,6 @@ def test_local_lda(self):
             self.assertEqual(str(model), str(model2))
 
     def test_distributed_lda(self):
-        if is_remote():
-            self.skipTest("Do not support Spark Connect.")
         spark = self.spark
         df = (
             spark.createDataFrame(