diff --git a/gymnasium/envs/mujoco/ant_v5.py b/gymnasium/envs/mujoco/ant_v5.py
index 5f15ad2cd..8b3ab177e 100644
--- a/gymnasium/envs/mujoco/ant_v5.py
+++ b/gymnasium/envs/mujoco/ant_v5.py
@@ -376,6 +376,23 @@ def step(self, action):
         xy_velocity = (xy_position_after - xy_position_before) / self.dt
         x_velocity, y_velocity = xy_velocity
 
+        observation = self._get_obs()
+        reward, reward_info = self._get_rew(x_velocity, action)
+        terminated = self.terminated
+        info = {
+            "x_position": self.data.qpos[0],
+            "y_position": self.data.qpos[1],
+            "distance_from_origin": np.linalg.norm(self.data.qpos[0:2], ord=2),
+            "x_velocity": x_velocity,
+            "y_velocity": y_velocity,
+            **reward_info,
+        }
+
+        if self.render_mode == "human":
+            self.render()
+        return observation, reward, terminated, False, info
+
+    def _get_rew(self, x_velocity: float, action):
         forward_reward = x_velocity * self._forward_reward_weight
         healthy_reward = self.healthy_reward
         rewards = forward_reward + healthy_reward
@@ -384,24 +401,16 @@ def step(self, action):
         contact_cost = self.contact_cost
         costs = ctrl_cost + contact_cost
 
-        observation = self._get_obs()
         reward = rewards - costs
-        terminated = self.terminated
-        info = {
+
+        reward_info = {
             "reward_forward": forward_reward,
             "reward_ctrl": -ctrl_cost,
             "reward_contact": -contact_cost,
             "reward_survive": healthy_reward,
-            "x_position": self.data.qpos[0],
-            "y_position": self.data.qpos[1],
-            "distance_from_origin": np.linalg.norm(self.data.qpos[0:2], ord=2),
-            "x_velocity": x_velocity,
-            "y_velocity": y_velocity,
         }
 
-        if self.render_mode == "human":
-            self.render()
-        return observation, reward, terminated, False, info
+        return reward, reward_info
 
     def _get_obs(self):
         position = self.data.qpos.flatten()
diff --git a/gymnasium/envs/mujoco/half_cheetah_v5.py b/gymnasium/envs/mujoco/half_cheetah_v5.py
index 6499ca7fb..0649f4503 100644
--- a/gymnasium/envs/mujoco/half_cheetah_v5.py
+++ b/gymnasium/envs/mujoco/half_cheetah_v5.py
@@ -245,22 +245,25 @@ def step(self, action):
         x_position_after = self.data.qpos[0]
         x_velocity = (x_position_after - x_position_before) / self.dt
 
-        ctrl_cost = self.control_cost(action)
+        observation = self._get_obs()
+        reward, reward_info = self._get_rew(x_velocity, action)
+        info = {"x_position": x_position_after, "x_velocity": x_velocity, **reward_info}
 
+        if self.render_mode == "human":
+            self.render()
+        return observation, reward, False, False, info
+
+    def _get_rew(self, x_velocity: float, action):
         forward_reward = self._forward_reward_weight * x_velocity
+        ctrl_cost = self.control_cost(action)
 
-        observation = self._get_obs()
         reward = forward_reward - ctrl_cost
-        info = {
-            "x_position": x_position_after,
-            "x_velocity": x_velocity,
+
+        reward_info = {
             "reward_forward": forward_reward,
             "reward_ctrl": -ctrl_cost,
         }
-
-        if self.render_mode == "human":
-            self.render()
-        return observation, reward, False, False, info
+        return reward, reward_info
 
     def _get_obs(self):
         position = self.data.qpos.flatten()
diff --git a/gymnasium/envs/mujoco/hopper_v5.py b/gymnasium/envs/mujoco/hopper_v5.py
index a1f1086fc..68813d190 100644
--- a/gymnasium/envs/mujoco/hopper_v5.py
+++ b/gymnasium/envs/mujoco/hopper_v5.py
@@ -316,29 +316,37 @@ def step(self, action):
         x_position_after = self.data.qpos[0]
         x_velocity = (x_position_after - x_position_before) / self.dt
 
-        ctrl_cost = self.control_cost(action)
+        observation = self._get_obs()
+        reward, reward_info = self._get_rew(x_velocity, action)
+        terminated = self.terminated
+        info = {
+            "x_position": x_position_after,
+            "z_distance_from_origin": self.data.qpos[1] - self.init_qpos[1],
+            "x_velocity": x_velocity,
+            **reward_info,
+        }
+
+        if self.render_mode == "human":
+            self.render()
+        return observation, reward, terminated, False, info
 
+    def _get_rew(self, x_velocity: float, action):
         forward_reward = self._forward_reward_weight * x_velocity
         healthy_reward = self.healthy_reward
-
         rewards = forward_reward + healthy_reward
+
+        ctrl_cost = self.control_cost(action)
         costs = ctrl_cost
 
-        observation = self._get_obs()
         reward = rewards - costs
-        terminated = self.terminated
-        info = {
+
+        reward_info = {
             "reward_forward": forward_reward,
             "reward_ctrl": -ctrl_cost,
             "reward_survive": healthy_reward,
-            "x_position": x_position_after,
-            "z_distance_from_origin": self.data.qpos[1] - self.init_qpos[1],
-            "x_velocity": x_velocity,
         }
 
-        if self.render_mode == "human":
-            self.render()
-        return observation, reward, terminated, False, info
+        return reward, reward_info
 
     def reset_model(self):
         noise_low = -self._reset_noise_scale
diff --git a/gymnasium/envs/mujoco/humanoid_v5.py b/gymnasium/envs/mujoco/humanoid_v5.py
index b0f078ec0..1834d6d48 100644
--- a/gymnasium/envs/mujoco/humanoid_v5.py
+++ b/gymnasium/envs/mujoco/humanoid_v5.py
@@ -497,23 +497,10 @@ def step(self, action):
         xy_velocity = (xy_position_after - xy_position_before) / self.dt
         x_velocity, y_velocity = xy_velocity
 
-        ctrl_cost = self.control_cost(action)
-        contact_cost = self.contact_cost
-        costs = ctrl_cost + contact_cost
-
-        forward_reward = self._forward_reward_weight * x_velocity
-        healthy_reward = self.healthy_reward
-
-        rewards = forward_reward + healthy_reward
-
         observation = self._get_obs()
-        reward = rewards - costs
+        reward, reward_info = self._get_rew(x_velocity, action)
         terminated = self.terminated
         info = {
-            "reward_survive": healthy_reward,
-            "reward_forward": forward_reward,
-            "reward_ctrl": -ctrl_cost,
-            "reward_contact": -contact_cost,
             "x_position": self.data.qpos[0],
             "y_position": self.data.qpos[1],
             "tendon_lenght": self.data.ten_length,
@@ -521,12 +508,33 @@ def step(self, action):
             "distance_from_origin": np.linalg.norm(self.data.qpos[0:2], ord=2),
             "x_velocity": x_velocity,
             "y_velocity": y_velocity,
+            **reward_info,
         }
 
         if self.render_mode == "human":
             self.render()
         return observation, reward, terminated, False, info
 
+    def _get_rew(self, x_velocity: float, action):
+        forward_reward = self._forward_reward_weight * x_velocity
+        healthy_reward = self.healthy_reward
+        rewards = forward_reward + healthy_reward
+
+        ctrl_cost = self.control_cost(action)
+        contact_cost = self.contact_cost
+        costs = ctrl_cost + contact_cost
+
+        reward = rewards - costs
+
+        reward_info = {
+            "reward_survive": healthy_reward,
+            "reward_forward": forward_reward,
+            "reward_ctrl": -ctrl_cost,
+            "reward_contact": -contact_cost,
+        }
+
+        return reward, reward_info
+
     def reset_model(self):
         noise_low = -self._reset_noise_scale
         noise_high = self._reset_noise_scale
diff --git a/gymnasium/envs/mujoco/humanoidstandup_v5.py b/gymnasium/envs/mujoco/humanoidstandup_v5.py
index 46e591ac8..99b35cc50 100644
--- a/gymnasium/envs/mujoco/humanoidstandup_v5.py
+++ b/gymnasium/envs/mujoco/humanoidstandup_v5.py
@@ -444,6 +444,21 @@ def step(self, action):
         self.do_simulation(action, self.frame_skip)
         pos_after = self.data.qpos[2]
 
+        reward, reward_info = self._get_rew(pos_after, action)
+        info = {
+            "x_position": self.data.qpos[0],
+            "y_position": self.data.qpos[1],
+            "z_distance_from_origin": self.data.qpos[2] - self.init_qpos[2],
+            "tendon_lenght": self.data.ten_length,
+            "tendon_velocity": self.data.ten_velocity,
+            **reward_info,
+        }
+
+        if self.render_mode == "human":
+            self.render()
+        return self._get_obs(), reward, False, False, info
+
+    def _get_rew(self, pos_after: float, action):
         uph_cost = (pos_after - 0) / self.model.opt.timestep
 
         quad_ctrl_cost = self._ctrl_cost_weight * np.square(self.data.ctrl).sum()
@@ -455,20 +470,14 @@ def step(self, action):
         quad_impact_cost = np.clip(quad_impact_cost, min_impact_cost, max_impact_cost)
 
         reward = uph_cost - quad_ctrl_cost - quad_impact_cost + 1
-        info = {
+
+        reward_info = {
             "reward_linup": uph_cost,
             "reward_quadctrl": -quad_ctrl_cost,
             "reward_impact": -quad_impact_cost,
-            "x_position": self.data.qpos[0],
-            "y_position": self.data.qpos[1],
-            "z_distance_from_origin": self.data.qpos[2] - self.init_qpos[2],
-            "tendon_lenght": self.data.ten_length,
-            "tendon_velocity": self.data.ten_velocity,
         }
 
-        if self.render_mode == "human":
-            self.render()
-        return self._get_obs(), reward, False, False, info
+        return reward, reward_info
 
     def reset_model(self):
         noise_low = -self._reset_noise_scale
diff --git a/gymnasium/envs/mujoco/inverted_double_pendulum_v5.py b/gymnasium/envs/mujoco/inverted_double_pendulum_v5.py
index dfa190ca2..bf2835577 100644
--- a/gymnasium/envs/mujoco/inverted_double_pendulum_v5.py
+++ b/gymnasium/envs/mujoco/inverted_double_pendulum_v5.py
@@ -194,27 +194,32 @@ def __init__(
     def step(self, action):
         self.do_simulation(action, self.frame_skip)
 
+        x, _, y = self.data.site_xpos[0]
         observation = self._get_obs()
+        terminated = bool(y <= 1)
+        reward, reward_info = self._get_rew(x, y, terminated)
 
-        x, _, y = self.data.site_xpos[0]
-        v1, v2 = self.data.qvel[1:3]
+        info = reward_info
 
-        terminated = bool(y <= 1)
+        if self.render_mode == "human":
+            self.render()
+        return observation, reward, terminated, False, info
 
+    def _get_rew(self, x, y, terminated):
+        v1, v2 = self.data.qvel[1:3]
         dist_penalty = 0.01 * x**2 + (y - 2) ** 2
         vel_penalty = 1e-3 * v1**2 + 5e-3 * v2**2
         alive_bonus = self._healthy_reward * int(not terminated)
+
         reward = alive_bonus - dist_penalty - vel_penalty
 
-        info = {
+        reward_info = {
             "reward_survive": alive_bonus,
             "distance_penalty": -dist_penalty,
             "velocity_penalty": -vel_penalty,
         }
 
-        if self.render_mode == "human":
-            self.render()
-        return observation, reward, terminated, False, info
+        return reward, reward_info
 
     def _get_obs(self):
         return np.concatenate(
diff --git a/gymnasium/envs/mujoco/pusher_v5.py b/gymnasium/envs/mujoco/pusher_v5.py
index 99d4eabf0..490c4b016 100644
--- a/gymnasium/envs/mujoco/pusher_v5.py
+++ b/gymnasium/envs/mujoco/pusher_v5.py
@@ -220,6 +220,16 @@ def __init__(
         }
 
     def step(self, action):
+        reward, reward_info = self._get_rew(action)
+        self.do_simulation(action, self.frame_skip)
+
+        observation = self._get_obs()
+        info = reward_info
+        if self.render_mode == "human":
+            self.render()
+        return observation, reward, False, False, info
+
+    def _get_rew(self, action):
         vec_1 = self.get_body_com("object") - self.get_body_com("tips_arm")
         vec_2 = self.get_body_com("object") - self.get_body_com("goal")
 
@@ -227,18 +237,15 @@ def step(self, action):
         reward_dist = -np.linalg.norm(vec_2) * self._reward_dist_weight
         reward_ctrl = -np.square(action).sum() * self._reward_control_weight
 
-        self.do_simulation(action, self.frame_skip)
-
-        observation = self._get_obs()
         reward = reward_dist + reward_ctrl + reward_near
-        info = {
+
+        reward_info = {
             "reward_dist": reward_dist,
             "reward_ctrl": reward_ctrl,
             "reward_near": reward_near,
         }
-        if self.render_mode == "human":
-            self.render()
-        return observation, reward, False, False, info
+
+        return reward, reward_info
 
     def reset_model(self):
         qpos = self.init_qpos
diff --git a/gymnasium/envs/mujoco/reacher_v5.py b/gymnasium/envs/mujoco/reacher_v5.py
index e8ba867d1..db23e1961 100644
--- a/gymnasium/envs/mujoco/reacher_v5.py
+++ b/gymnasium/envs/mujoco/reacher_v5.py
@@ -197,21 +197,28 @@ def __init__(
         }
 
     def step(self, action):
+        reward, reward_info = self._get_rew(action)
+        self.do_simulation(action, self.frame_skip)
+
+        observation = self._get_obs()
+        info = reward_info
+        if self.render_mode == "human":
+            self.render()
+        return observation, reward, False, False, info
+
+    def _get_rew(self, action):
         vec = self.get_body_com("fingertip") - self.get_body_com("target")
         reward_dist = -np.linalg.norm(vec) * self._reward_dist_weight
         reward_ctrl = -np.square(action).sum() * self._reward_control_weight
 
-        self.do_simulation(action, self.frame_skip)
-
-        observation = self._get_obs()
         reward = reward_dist + reward_ctrl
-        info = {
+
+        reward_info = {
             "reward_dist": reward_dist,
             "reward_ctrl": reward_ctrl,
         }
-        if self.render_mode == "human":
-            self.render()
-        return observation, reward, False, False, info
+
+        return reward, reward_info
 
     def reset_model(self):
         qpos = (
diff --git a/gymnasium/envs/mujoco/swimmer_v5.py b/gymnasium/envs/mujoco/swimmer_v5.py
index c49267bce..a231cc627 100644
--- a/gymnasium/envs/mujoco/swimmer_v5.py
+++ b/gymnasium/envs/mujoco/swimmer_v5.py
@@ -234,20 +234,15 @@ def step(self, action):
         xy_velocity = (xy_position_after - xy_position_before) / self.dt
         x_velocity, y_velocity = xy_velocity
 
-        forward_reward = self._forward_reward_weight * x_velocity
-
-        ctrl_cost = self.control_cost(action)
-
         observation = self._get_obs()
-        reward = forward_reward - ctrl_cost
+        reward, reward_info = self._get_rew(x_velocity, action)
         info = {
-            "reward_forward": forward_reward,
-            "reward_ctrl": -ctrl_cost,
             "x_position": xy_position_after[0],
             "y_position": xy_position_after[1],
             "distance_from_origin": np.linalg.norm(xy_position_after, ord=2),
             "x_velocity": x_velocity,
             "y_velocity": y_velocity,
+            **reward_info,
         }
 
         if self.render_mode == "human":
@@ -255,6 +250,19 @@ def step(self, action):
 
         return observation, reward, False, False, info
 
+    def _get_rew(self, x_velocity: float, action):
+        forward_reward = self._forward_reward_weight * x_velocity
+        ctrl_cost = self.control_cost(action)
+
+        reward = forward_reward - ctrl_cost
+
+        reward_info = {
+            "reward_forward": forward_reward,
+            "reward_ctrl": -ctrl_cost,
+        }
+
+        return reward, reward_info
+
     def _get_obs(self):
         position = self.data.qpos.flatten()
         velocity = self.data.qvel.flatten()
diff --git a/gymnasium/envs/mujoco/walker2d_v5.py b/gymnasium/envs/mujoco/walker2d_v5.py
index dde3e9b03..555ca4944 100644
--- a/gymnasium/envs/mujoco/walker2d_v5.py
+++ b/gymnasium/envs/mujoco/walker2d_v5.py
@@ -310,24 +310,14 @@ def step(self, action):
         x_position_after = self.data.qpos[0]
         x_velocity = (x_position_after - x_position_before) / self.dt
 
-        ctrl_cost = self.control_cost(action)
-
-        forward_reward = self._forward_reward_weight * x_velocity
-        healthy_reward = self.healthy_reward
-
-        rewards = forward_reward + healthy_reward
-        costs = ctrl_cost
-
         observation = self._get_obs()
-        reward = rewards - costs
+        reward, reward_info = self._get_rew(x_velocity, action)
         terminated = self.terminated
         info = {
-            "reward_forward": forward_reward,
-            "reward_ctrl": -ctrl_cost,
-            "reward_survive": healthy_reward,
             "x_position": x_position_after,
             "z_distance_from_origin": self.data.qpos[1] - self.init_qpos[1],
             "x_velocity": x_velocity,
+            **reward_info,
         }
 
         if self.render_mode == "human":
@@ -335,6 +325,23 @@ def step(self, action):
 
         return observation, reward, terminated, False, info
 
+    def _get_rew(self, x_velocity: float, action):
+        forward_reward = self._forward_reward_weight * x_velocity
+        healthy_reward = self.healthy_reward
+        rewards = forward_reward + healthy_reward
+
+        ctrl_cost = self.control_cost(action)
+        costs = ctrl_cost
+        reward = rewards - costs
+
+        reward_info = {
+            "reward_forward": forward_reward,
+            "reward_ctrl": -ctrl_cost,
+            "reward_survive": healthy_reward,
+        }
+
+        return reward, reward_info
+
     def reset_model(self):
         noise_low = -self._reset_noise_scale
         noise_high = self._reset_noise_scale