Extant research finds inconclusive evidence about the CEO horizon problem. One possibility is that compensation committees design CEO compensation in a way that discourages retiring CEOs from opportunistic earnings management and R&D reduction. However, compensation committees dominated by co-opted directors may not be as effective as those with fewer co-opted directors in mitigating the CEO horizon problem, because directors co-opted by the CEO tend to bias their decisions in favor of the CEO. I find that compensation committees dominated by co-opted directors are associated with higher CEO compensation packages. I document R&D reduction and accruals management in firms with retiring CEOs and compensation committees dominated by co-opted directors, and find that R&D reduction and income-increasing accruals are less discouraged by compensation committees dominated by co-opted directors when deciding CEO compensation. I also examine the effect of boards of directors and compensation committee characteristics on CEO compensation and on mitigating the CEO horizon problem. I find that CEO compensation positively associates with CEO power, director independence, and the percentage of busy directors, and negatively associates with board of directors and committee size and director ownership. Moreover, I find that retiring CEOs are more likely to reduce R&D expenditures when CEOs have more power, and director tenure is longer; retiring CEOs in firms with large boards of directors and compensation committees are less likely to manage accruals.