diff --git a/cdisc_rules_engine/models/actions.py b/cdisc_rules_engine/models/actions.py
index bb81c274e..63682ee4c 100644
--- a/cdisc_rules_engine/models/actions.py
+++ b/cdisc_rules_engine/models/actions.py
@@ -191,6 +191,10 @@ def generate_targeted_error_object(  # noqa: C901
                     targets,
                 )
 
+            grouping_variables = [
+                x.replace("--", self.dataset_metadata.wildcard_replacement or "")
+                for x in grouping_variables
+            ]
             missing_grouping_vars = [
                 var for var in grouping_variables if var not in data.columns
             ]
diff --git a/tests/unit/test_actions.py b/tests/unit/test_actions.py
index 81c81ce28..f025849f7 100644
--- a/tests/unit/test_actions.py
+++ b/tests/unit/test_actions.py
@@ -105,11 +105,159 @@ def test_targeted_error_object_with_dataset_sensitivity():
     assert error["value"] == {"TEST": 1, "MISSING": "Not in dataset"}
 
 
-def test_targeted_error_object_with_group_sensitivity():
+@pytest.mark.parametrize(
+    "grouping_variables, expected_column, df, expected_groups",
+    [
+        # Case 1: explicit column name, no wildcard
+        (
+            ["SETCD"],
+            "SETCD",
+            pd.DataFrame(
+                {
+                    "TXPARMCD": [
+                        "PLANSUB",
+                        "PLANSUB",
+                        "PLANSUB",
+                        "OTHER",
+                        "OTHER",
+                        "OTHER",
+                        "OTHER",
+                        "VALUE",
+                        "VALUE",
+                    ],
+                    "SETCD": [
+                        "SET1",
+                        "SET1",
+                        "SET1",
+                        "SET2",
+                        "SET2",
+                        "SET2",
+                        "SET2",
+                        "SET3",
+                        "SET3",
+                    ],
+                    "USUBJID": [
+                        "001",
+                        "001",
+                        "001",
+                        "002",
+                        "002",
+                        "002",
+                        "002",
+                        "003",
+                        "003",
+                    ],
+                }
+            ),
+            ["SET1", "SET2", "SET3"],
+        ),
+        # Case 2: wildcard "--PARMCD" → "TXPARMCD", grouped only on TXPARMCD
+        # Each distinct TXPARMCD value should produce one value
+        (
+            ["--PARMCD"],
+            "TXPARMCD",
+            pd.DataFrame(
+                {
+                    "TXPARMCD": [
+                        "PLANSUB",
+                        "PLANSUB",
+                        "PLANSUB",
+                        "OTHER",
+                        "OTHER",
+                        "OTHER",
+                        "OTHER",
+                        "VALUE",
+                        "VALUE",
+                    ],
+                    "SETCD": [
+                        "SET1",
+                        "SET1",
+                        "SET1",
+                        "SET2",
+                        "SET2",
+                        "SET2",
+                        "SET2",
+                        "SET3",
+                        "SET3",
+                    ],
+                    "USUBJID": [
+                        "001",
+                        "001",
+                        "001",
+                        "002",
+                        "002",
+                        "002",
+                        "002",
+                        "003",
+                        "003",
+                    ],
+                }
+            ),
+            ["PLANSUB", "OTHER", "VALUE"],
+        ),
+        # Case 3: grouped on both "--PARMCD" (→ "TXPARMCD") and "SETCD"
+        # TXPARMCD varies within SETCD groups to prove both columns contribute to grouping:
+        #   SET1 has PLANSUB + OTHER  → 2 groups
+        #   SET2 has OTHER only       → 1 group
+        #   SET3 has VALUE + PLANSUB  → 2 groups
+        # Total: 5 distinct (TXPARMCD, SETCD) combinations
+        (
+            ["--PARMCD", "SETCD"],
+            "SETCD",
+            pd.DataFrame(
+                {
+                    "TXPARMCD": [
+                        "PLANSUB",
+                        "PLANSUB",
+                        "OTHER",
+                        "OTHER",
+                        "OTHER",
+                        "OTHER",
+                        "OTHER",
+                        "VALUE",
+                        "VALUE",
+                        "PLANSUB",
+                        "PLANSUB",
+                    ],
+                    "SETCD": [
+                        "SET1",
+                        "SET1",
+                        "SET1",
+                        "SET1",
+                        "SET2",
+                        "SET2",
+                        "SET2",
+                        "SET3",
+                        "SET3",
+                        "SET3",
+                        "SET3",
+                    ],
+                    "USUBJID": [
+                        "001",
+                        "001",
+                        "001",
+                        "001",
+                        "002",
+                        "002",
+                        "002",
+                        "003",
+                        "003",
+                        "003",
+                        "003",
+                    ],
+                }
+            ),
+            ["SET1", "SET1", "SET2", "SET3", "SET3"],  # two SET1s and two SET3s
+        ),
+    ],
+)
+def test_targeted_error_object_with_group_sensitivity(
+    grouping_variables, expected_column, df, expected_groups
+):
     dummy_rule = {
         "core_id": "FB4607",
         "sensitivity": "Group",
-        "grouping_variables": ["SETCD"],
+        "grouping_variables": grouping_variables,
         "conditions": {
             "all": [
                 {
@@ -128,35 +276,6 @@ def test_targeted_error_object_with_group_sensitivity():
         "output_variables": ["SETCD", "TXPARMCD"],
     }
 
-    # Create test data with 3 groups (SETCD), multiple records each
-    df = pd.DataFrame.from_dict(
-        {
-            "TXPARMCD": [
-                "PLANSUB",
-                "PLANSUB",
-                "PLANSUB",
-                "OTHER",
-                "OTHER",
-                "OTHER",
-                "OTHER",
-                "VALUE",
-                "VALUE",
-            ],
-            "SETCD": [
-                "SET1",
-                "SET1",
-                "SET1",
-                "SET2",
-                "SET2",
-                "SET2",
-                "SET2",
-                "SET3",
-                "SET3",
-            ],
-            "USUBJID": ["001", "001", "001", "002", "002", "002", "002", "003", "003"],
-        }
-    )
-
     variable = DatasetVariable(PandasDataset(df))
     dataset_metadata = SDTMDatasetMetadata(
         first_record={"DOMAIN": "TX"}, filename="tx.xpt"
@@ -168,21 +287,18 @@ def test_targeted_error_object_with_group_sensitivity():
         targets, df, "Missing required TXPARMCD value per set"
     )
 
-    # Should have exactly 3 errors (one per SETCD group)
-    assert len(result.errors) == 3
+    assert len(result.errors) == len(expected_groups)
 
-    # Extract SETCD values from errors
-    setcd_values = [
-        error.to_representation()["value"]["SETCD"] for error in result.errors
-    ]
-    assert sorted(setcd_values) == ["SET1", "SET2", "SET3"]
+    actual_values = sorted(
+        error.to_representation()["value"][expected_column] for error in result.errors
+    )
+    assert actual_values == sorted(expected_groups)
 
     # Each error should have row information from the first record in its group
     for error in result.errors:
         error_repr = error.to_representation()
         assert "row" in error_repr
-        assert "SETCD" in error_repr["value"]
-        assert "TXPARMCD" in error_repr["value"]
+        assert expected_column in error_repr["value"]
 
 
 def test_group_sensitivity_missing_grouping_variables():