diff --git a/python/tvm/relay/frontend/onnx.py b/python/tvm/relay/frontend/onnx.py
index 997aa6240..abfa5629d 100644
--- a/python/tvm/relay/frontend/onnx.py
+++ b/python/tvm/relay/frontend/onnx.py
@@ -5061,10 +5061,27 @@ class Momentum(OnnxOpConverter):
         return _expr.TupleWrapper(_expr.Tuple(result), len(result))
 
 
+class Round(OnnxOpConverter):
+    """Operator converter for round op."""
+
+    @classmethod
+    def _impl_v11(cls, inputs, attr, params):
+        # Onnx round uses Banker's rounding which rounds .5 to the nearest even integer
+
+        x = inputs[0]
+        half = _expr.const(0.5, dtype="float32")
+        one = _expr.const(1, dtype="float32")
+        two = _expr.const(2, dtype="float32")
+
+        rounded = _op.ceil(x - half)
+        bankers_mask = one - (_op.ceil(x + half) - _op.floor(x + half))
+        non_even = _op.abs(_op.mod(rounded, two))
+        return rounded + (bankers_mask * non_even)
+
+
 # compatible operators that do NOT require any conversion.
 _identity_list = []
 
-
 # _convert_map defines maps of name to converter functor(callable)
 # for 1 to 1 mapping, use Renamer if nothing but name is different
 # use AttrCvt if attributes need to be converted
@@ -5109,7 +5126,7 @@ def _get_convert_map(opset):
         "Reciprocal": Reciprocal.get_converter(opset),
         "Floor": Renamer("floor"),
         "Ceil": Renamer("ceil"),
-        "Round": Renamer("round"),
+        "Round": Round.get_converter(opset),
         "IsInf": IsInf.get_converter(opset),
         "IsNaN": Renamer("isnan"),
         "Sqrt": Renamer("sqrt"),
diff --git a/tests/python/frontend/onnx/test_forward.py b/tests/python/frontend/onnx/test_forward.py
index c4cd93aa7..ebaad9b4c 100644
--- a/tests/python/frontend/onnx/test_forward.py
+++ b/tests/python/frontend/onnx/test_forward.py
@@ -5183,7 +5183,6 @@ unsupported_onnx_tests = [
     "test_reduce_sum_negative_axes_keepdims_example",
     "test_reduce_sum_negative_axes_keepdims_random",
     "test_rnn_seq_length",
-    "test_round",
     "test_sequence_insert_at_back",
     "test_sequence_insert_at_front",
     "test_simple_rnn_batchwise",