diff --git a/src/ezmsg/sigproc/butterworthfilter.py b/src/ezmsg/sigproc/butterworthfilter.py
index 942f5dc..c3bcb67 100644
--- a/src/ezmsg/sigproc/butterworthfilter.py
+++ b/src/ezmsg/sigproc/butterworthfilter.py
@@ -97,8 +97,13 @@ def butter_design_fun(
 
     """
     coefs = None
-    if order > 0:
-        btype, cutoffs = ButterworthFilterSettings(order=order, cuton=cuton, cutoff=cutoff).filter_specs()
+    # order > 0 with neither corner set is a valid "no filter" request:
+    # filter_specs() returns None, which we treat as passthrough (coefs stays
+    # None -> the transformer returns the input unchanged) rather than crashing
+    # on an unpack of None.
+    specs = ButterworthFilterSettings(order=order, cuton=cuton, cutoff=cutoff).filter_specs()
+    if order > 0 and specs is not None:
+        btype, cutoffs = specs
         coefs = scipy.signal.butter(
             order,
             Wn=cutoffs,
diff --git a/tests/unit/test_butter.py b/tests/unit/test_butter.py
index e8f0db6..44fe257 100644
--- a/tests/unit/test_butter.py
+++ b/tests/unit/test_butter.py
@@ -288,6 +288,26 @@ def test_butterworth_empty_order0_passthrough():
     check_empty_result(result)
 
 
+def test_butterworth_order_no_corners_passthrough():
+    """order > 0 with neither cuton nor cutoff is a valid "no filter" request.
+
+    Regression test: this used to crash in butter_design_fun on an unpack of
+    None from filter_specs(); now it designs no filter and passes data through.
+    """
+    from ezmsg.sigproc.butterworthfilter import (
+        ButterworthFilterSettings,
+        ButterworthFilterTransformer,
+        butter_design_fun,
+    )
+
+    assert butter_design_fun(fs=1000.0, order=4) is None
+
+    proc = ButterworthFilterTransformer(ButterworthFilterSettings(order=4, axis="time"))
+    msg = make_msg()
+    result = proc(msg)
+    assert np.array_equal(result.data, msg.data)
+
+
 def test_butterworth_empty_first():
     from ezmsg.sigproc.butterworthfilter import ButterworthFilterSettings, ButterworthFilterTransformer