A large body of the corporate governance literature examines the disciplinary role of outside directors in overseeing the CEO. Although it is certainly a critical factor in effective monitoring, independence alone is not sufficient. Fulfilling the monitoring role also requires a skilled and knowledgeable board (Acharya, Myers, & Rajan, 2011; Adams & Ferreira 2007; Raheja, 2005). The skills and knowledge needed for monitoring vary with the type of CEO activity being monitored. For certain managerial actions that require sufficient firm-specific knowledge and expertise to exercise discipline, board informedness could be at least as critical as board independence. Given the trade-off between informedness and independence, outside directors are not necessarily better monitors than inside directors due to information disadvantages.1 In this study, we examine whether and to what extent an independent board constrains the CEO from taking real actions to manage financial performance. We refer to real earnings management as the CEO’s purposeful intervention in normal business practice in an effort to influence the output of the accounting system (Gunny, 2010; Roychowdhury, 2006).2 Relative to accrual-based earnings management, real earnings management is inherently more difficult to detect and requires more firm-specific information to understand because it can involve any real decision that deviates from normal business practice (Cohen, Dey, & Lys, 2008; Lo, 2008). Lo (2008) compares the two earnings management methods and concludes that managers are willing to engage in real earnings management that is costly to the firm because such actions are harder to detect; with the uncertainty inherent in business environments, there is no benchmark to determine what should have been done under any particular situation . . . (p. 353) Therefore, without sufficient firm-specific knowledge, and thus the expertise to identify deviations from normal business practice, outside directors could be ineffective at detecting real earnings management. A recent study by Armstrong, Core, and Guay (2014) highlights the informational demands of an independent board to perform its monitoring duties by showing that an increase in board independence results in an increase in firm transparency. They interpret their results to indicate that outside directors require transparency to perform the monitoring duties, and that firms can change their corporate transparency to suit the informational demands of a particular board structure. Although it is an implicit assumption in Armstrong et al. (2014) that board informedness plays a critical role in effective monitoring, we provide empirical evidence on its importance and trade-off against board independence in the context of monitoring real earnings management. Relative to their inside counterparts, outside directors are often less knowledgeable with regard to the day-to-day operations of the firm. In practice, outside directors must largely rely on management to provide them with the information necessary to perform their monitoring duties. Given that timely information transfer across individuals is costly and outside directors are typically busy individuals who have many demands on their time, they may not invest the time and effort necessary to become as informed as the management. To effectively constrain real earnings management, the board likely faces a pronounced trade-off between informedness and independence. It’s not clear, ex ante, whether the welldocumented result that the greater the board independence, the less the accrual-based earnings management (e.g., Klein, 2002; Peasnell, Pope, & Young, 2005) should also apply to the real earnings management context. Our first research objective is to examine whether, in general, board independence is associated with real earnings management. A negative relation would indicate that board independence plays a dominant role in monitoring of real earnings management. By contrast, an insignificant (or even positive) association would suggest that board independence alone is unlikely to be sufficient for effective monitoring of real earnings management.