Despite inconclusive empirical evidence on the impact on firm performance of having a board chair who is not the CEO, there is a growing incidence of US public companies appointing board chairs who are not current or former executives. Using a large panel dataset from 1996 to 2005, we study the determinants and impact of outside chairs on firm performance. We find that having an outside chair positively and significantly impacts firm performance. This finding is robust to several tests that control for potential endogeneity and reverse causality interpretations. Consistent with recent literature on optimal board structures, we find that the costs and benefits of having an outside chair depend on firm characteristics. Specifically, we find that having an outside chair is more valuable when the CEO has greater bargaining power or when monitoring costs to external investors are high. On the other hand, having an outside chair is less valuable in operationally complex firms. Finally, we document that shareholders react favorably to announcements of outside chair appointments.