diff --git a/.jenkins/pytorch/test.sh b/.jenkins/pytorch/test.sh
index 1e4a9a52066..a74a3e81580 100755
--- a/.jenkins/pytorch/test.sh
+++ b/.jenkins/pytorch/test.sh
@@ -450,6 +450,8 @@ test_xla() {
 # nightly version.
 test_forward_backward_compatibility() {
   set -x
+  # create a dummy ts model at this version
+  python test/create_dummy_torchscript_model.py /tmp/model_new.pt
   pushd test/forward_backward_compatibility
   python -m venv venv
   # shellcheck disable=SC1091
@@ -457,10 +459,21 @@ test_forward_backward_compatibility() {
   pip_install --pre torch -f https://download.pytorch.org/whl/nightly/cpu/torch_nightly.html
   pip show torch
   python dump_all_function_schemas.py --filename nightly_schemas.txt
+  # FC: verify newmodel can be load with old code.
+  if ! python ../load_torchscript_model.py /tmp/model_new.pt; then
+      echo "FC check failed: new model cannot be load in old code"
+      return 1
+  fi
+  python ../create_dummy_torchscript_model.py /tmp/model_old.pt
   deactivate
   rm -r venv
   pip show torch
   python check_forward_backward_compatibility.py --existing-schemas nightly_schemas.txt
+  # BC: verify old model can be load with new code
+  if ! python ../load_torchscript_model.py /tmp/model_old.pt; then
+      echo "BC check failed: old model cannot be load in new code"
+      return 1
+  fi
   popd
   set +x
   assert_git_not_dirty
diff --git a/test/create_dummy_torchscript_model.py b/test/create_dummy_torchscript_model.py
new file mode 100644
index 00000000000..ffd869e27f0
--- /dev/null
+++ b/test/create_dummy_torchscript_model.py
@@ -0,0 +1,28 @@
+# Usage: python create_dummy_model.py <name_of_the_file>
+import sys
+import torch
+from torch import nn
+
+
+class NeuralNetwork(nn.Module):
+
+    def __init__(self):
+        super(NeuralNetwork, self).__init__()
+        self.flatten = nn.Flatten()
+        self.linear_relu_stack = nn.Sequential(
+            nn.Linear(28 * 28, 512),
+            nn.ReLU(),
+            nn.Linear(512, 512),
+            nn.ReLU(),
+            nn.Linear(512, 10),
+        )
+
+    def forward(self, x):
+        x = self.flatten(x)
+        logits = self.linear_relu_stack(x)
+        return logits
+
+
+if __name__ == '__main__':
+    jit_module = torch.jit.script(NeuralNetwork())
+    torch.jit.save(jit_module, sys.argv[1])
diff --git a/test/load_torchscript_model.py b/test/load_torchscript_model.py
new file mode 100644
index 00000000000..dc8d4159d7f
--- /dev/null
+++ b/test/load_torchscript_model.py
@@ -0,0 +1,6 @@
+import sys
+import torch
+
+if __name__ == '__main__':
+    print(torch.jit.load(sys.argv[1]))
+    sys.exit(0)