In this paper, I show that David Lewis' theoretical framework is basically sufficient to solve the problem of causal preemption. Causal preemption cases have been regarded as a counterexample to the counterfactual analysis of causation. This view is, however, incorrect. Even under the counterfactual analysis, the problem can be solved by adopting the event theory which defines events as properties of spatio-temporal regions. The source of the problem lies in the assumption that two distinct intuitions should be preserved in causal preemption cases. The appropriate event theory is required in order to sort out the intuitions.
