Abstract A significant and growing percentage of U.S. firms now have boards where the CEO is the only employee director (hereinafter fully independent boards). This paper studies whether and how this practice impacts board effectiveness. I find that fully independent boards are associated with a significant reduction in firm performance. Further tests suggest two channels for this effect. First, full independence deprives the board of spontaneous and regular access to the firm-specific information of other senior executives. Second, full independence eliminates the first-hand exposure of future CEOs to board-level discussions of strategy, which steepens the learning curve for eventually promoted candidates.