To mitigate the concern that companies with overconfident CEOs have company characteristics that are correlated with audit risk, we use propensity score matching to identify companies that are similar in characteristics but differ in the over- confidence of the CEO. Using a sample based on propensity score matching, we continue to find that, in the absence of a strong audit committee, managerial overconfidence is negatively associated with audit fees. These results suggest that dif- ferences in observable company characteristics are unlikely to be driving our results. Additionally, our findings are robust to the use of a treatment effects model to control for endogeneity and selection bias.