In my dissertation, I develop a theoretical model that explains how leaders' daily work demands and recovery affect their leadership behaviors. In a departure from the trait approach of leadership which suggests that leaders tend to behave in certain ways that are determined by their heritable characteristics such as personality and intelligence (e.g., Bono & Judge, 2002), and from the contingency approach that suggests leaders behave in ways that are most suitable to the situation based on the needs of followers and the demands of their tasks (e.g., House, 1971), this dissertation draws from the transactional theory of stress (Lazarus & Folkman, 1984) and positions the stressful demands that leaders experience at work as important determinants of their leadership behaviors. Specifically, I propose that leaders' daily challenge demands (e.g., workload, time pressure, responsibilities) are positively related to job engagement whereas their daily hindrance demands (e.g., role ambiguity, office politics, and hassles) are negatively related to engagement. Engagement, in turn, is positively related to transformational and transactional leadership and negatively related to laissez-faire leadership and abusive supervision. Meanwhile, both challenge and hindrance demands are positively related to strain, which is negatively related to transformational and transactional leadership, and is positively related to laissez-faire leadership and abusive supervision. In addition, leaders' daily after-work recovery experience influences the mediating roles of engagement and strain in the relationships between work demands and leadership behaviors. Specifically, daily recovery moderates both the first stage (i.e., the linkages between work demands and engagement and strain) and the second stage (i.e., the linkages between engagement and strain and leadership behaviors) of the mediation. I test this two-level dual-stage moderated mediation model using a two-week experience sampling design. The sample consists of 26 supervisors and 73 employees who directly report to these supervisors from the flood control district of a metropolitan county in the Southwest United States. Results suggest that leaders' daily challenge demands have a positive influence on transformational leadership attributable to engagement, a negative influence on abusive supervision attributable to engagement, and a positive influence on abusive supervision attributable to strain. Leaders' daily hindrance demands, in contrast, have a positive influence on abusive supervision attributable to strain. In addition, leaders' daily recovery moderates the relationship between strain and laissez-faire leadership so that hindrance demands have a positive influence on laissez-faire leadership when the individual is poorly recovered. Leaders' daily recovery also moderates the relationship between strain and abusive supervision so that hindrance demands have a stronger positive influence on abusive supervision through strain when the individual is poorly recovered.