diff --git a/pymc/distributions/continuous.py b/pymc/distributions/continuous.py
index 97ad4035e1..50f3f1f368 100644
--- a/pymc/distributions/continuous.py
+++ b/pymc/distributions/continuous.py
@@ -2924,6 +2924,12 @@ def dist(cls, mu=0.0, kappa=None, *args, **kwargs):
         assert_negative_support(kappa, "kappa", "VonMises")
         return super().dist([mu, kappa], *args, **kwargs)
 
+    def get_moment(rv, size, mu, kappa):
+        mu, _ = at.broadcast_arrays(mu, kappa)
+        if not rv_size_is_none(size):
+            mu = at.full(size, mu)
+        return mu
+
 
 class SkewNormalRV(RandomVariable):
     name = "skewnormal"
diff --git a/pymc/tests/test_distributions_moments.py b/pymc/tests/test_distributions_moments.py
index ad6e234305..b5d1b0cfd9 100644
--- a/pymc/tests/test_distributions_moments.py
+++ b/pymc/tests/test_distributions_moments.py
@@ -54,6 +54,7 @@
     Triangular,
     TruncatedNormal,
     Uniform,
+    VonMises,
     Wald,
     Weibull,
     ZeroInflatedBinomial,
@@ -438,6 +439,21 @@ def test_pareto_moment(alpha, m, size, expected):
     assert_moment_is_expected(model, expected)
 
 
+@pytest.mark.parametrize(
+    "mu, kappa, size, expected",
+    [
+        (0, 1, None, 0),
+        (0, np.ones(4), None, np.zeros(4)),
+        (np.arange(4), 0.5, None, np.arange(4)),
+        (np.arange(4), np.arange(1, 5), (2, 4), np.full((2, 4), np.arange(4))),
+    ],
+)
+def test_vonmises_moment(mu, kappa, size, expected):
+    with Model() as model:
+        VonMises("x", mu=mu, kappa=kappa, size=size)
+    assert_moment_is_expected(model, expected)
+
+
 @pytest.mark.parametrize(
     "mu, lam, phi, size, expected",
     [