diff --git a/tf_agents/bandits/policies/linear_bandit_policy.py b/tf_agents/bandits/policies/linear_bandit_policy.py
index 73e5af527..4f8f203f0 100644
--- a/tf_agents/bandits/policies/linear_bandit_policy.py
+++ b/tf_agents/bandits/policies/linear_bandit_policy.py
@@ -107,7 +107,7 @@ def __init__(
       num_samples: list of number of samples per arm, unless the policy accepts
         per-arm features, in which case this is just the number of samples seen.
       time_step_spec: A `TimeStep` spec of the expected time_steps.
-      exploration_strategy: An Enum of type ExplortionStrategy. The strategy
+      exploration_strategy: An Enum of type ExplorationStrategy. The strategy
         used for choosing the actions to incorporate exploration. Currently
         supported strategies are `optimistic` and `sampling`.
       alpha: a float value used to scale the confidence intervals.