diff --git a/tests/data_collector/callbacks/test_step_data_callback.py b/tests/data_collector/callbacks/test_step_data_callback.py
index 9dd06f35..f50efc24 100644
--- a/tests/data_collector/callbacks/test_step_data_callback.py
+++ b/tests/data_collector/callbacks/test_step_data_callback.py
@@ -111,8 +111,9 @@ def test_data_collector_step_data_callback(data_format, register_dummy_envs):
 
 
 @pytest.mark.parametrize("data_format", get_storage_keys())
+@pytest.mark.parametrize("infos_format", ["dict", "list"])
 def test_data_collector_step_data_callback_info_correction(
-    data_format, register_dummy_envs
+    data_format, register_dummy_envs, infos_format
 ):
     """Test DataCollector wrapper and Minari dataset creation."""
     dataset_id = "dummy-inconsistent-info-v0"
@@ -123,6 +124,7 @@ def test_data_collector_step_data_callback_info_correction(
         record_infos=True,
         step_data_callback=CustomSubsetInfoPadStepDataCallback,
         data_format=data_format,
+        infos_format=infos_format,
     )
     num_episodes = 10
 
@@ -164,7 +166,8 @@ def test_data_collector_step_data_callback_info_correction(
         record_infos=True,
     )
     # here we are checking to make sure that if we have an environment changing its info
-    # structure across steps, IT IS OK!
+    # structure across steps, it results in an error when infos_format is the default (dict)
+    # behaviour.
     with pytest.raises(ValueError):
         num_episodes = 10
         env.reset(seed=42)
@@ -176,5 +179,8 @@ def test_data_collector_step_data_callback_info_correction(
                 _, _, terminated, truncated, _ = env.step(action)
 
             env.reset()
-        raise ValueError("This should BE reached")
+        if infos_format == "list":
+            raise ValueError(
+                "This should be raised to indicate the test succeeded for infos_format='list'"
+            )
     env.close()