diff --git a/pytensor/tensor/signal/conv.py b/pytensor/tensor/signal/conv.py
index 1152f02d8a..ab2856b694 100644
--- a/pytensor/tensor/signal/conv.py
+++ b/pytensor/tensor/signal/conv.py
@@ -119,8 +119,8 @@ def convolve1d(
     if mode == "same":
         # We implement "same" as "valid" with padded `in1`.
         in1_batch_shape = tuple(in1.shape)[:-1]
-        zeros_left = in2.shape[0] // 2
-        zeros_right = (in2.shape[0] - 1) // 2
+        zeros_left = in2.shape[-1] // 2
+        zeros_right = (in2.shape[-1] - 1) // 2
         in1 = join(
             -1,
             zeros((*in1_batch_shape, zeros_left), dtype=in2.dtype),
diff --git a/tests/tensor/signal/test_conv.py b/tests/tensor/signal/test_conv.py
index 968e408485..fe353b18fb 100644
--- a/tests/tensor/signal/test_conv.py
+++ b/tests/tensor/signal/test_conv.py
@@ -47,3 +47,16 @@ def test_convolve1d_batch():
     res_np = np.convolve(x_test[0], y_test[0])
     np.testing.assert_allclose(res[0], res_np, rtol=rtol)
     np.testing.assert_allclose(res[1], res_np, rtol=rtol)
+
+
+def test_convolve1d_batch_same():
+    x = matrix("data")
+    y = matrix("kernel")
+    out = convolve1d(x, y, mode="same")
+
+    rng = np.random.default_rng(38)
+    x_test = rng.normal(size=(2, 8)).astype(x.dtype)
+    y_test = rng.normal(size=(2, 8)).astype(x.dtype)
+
+    res = out.eval({x: x_test, y: y_test})
+    assert res.shape == (2, 8)