diff --git a/examples/mujoco/train_ppo_gym.py b/examples/mujoco/train_ppo_gym.py
index 13d1b933f..73c918250 100644
--- a/examples/mujoco/train_ppo_gym.py
+++ b/examples/mujoco/train_ppo_gym.py
@@ -130,8 +130,7 @@ def make_env(test):
         return env
 
     sample_env = gym.make(args.env)
-    timestep_limit = sample_env.spec.tags.get(
-        'wrapper_config.TimeLimit.max_episode_steps')
+    timestep_limit = sample_env.spec.max_episode_steps
     obs_space = sample_env.observation_space
     action_space = sample_env.action_space