Based on Bandura's (1986) social cognitive theory, this dissertation examined the contextual effects of leaders' coaching style on followers' adaptation and specific emotions, and, in turn, their relationships with individual work outcomes. This two-part (i.e. two-study) doctoral thesis addressed largely unexplored contextual antecedents of individual adaptation. Specifically, I proposed that two styles of coaching, guidance and facilitation, have differentiating effects on individual adaptation, job related feelings of anxiety, and employee emotions (e.g. cheerfulness, dejection). I also explored learning-self efficacy as a mediator of the relationships between employee emotions and both task performance and adaptive task performance; and adaptive behavior and job-related feelings of anxiety as mediators of the relationships between coaching style and work outcomes (i.e. task performance, creativity, emotional exhaustion). Study 1 is a field study with 371 employees and 49 immediate supervisors. Here I tested a model in which individual adaptive behavior and job-related feelings of anxiety mediate the relationships between supervisor's coaching style and three individual work outcomes, emotional exhaustion, task performance and creativity. Two coaching styles (guidance and facilitation coaching) were significantly related to both adaptive behavior and job-related feelings of anxiety, but in opposite directions. Moreover, there were contrasting effects of adaptive behavior and job-related feelings of anxiety on emotional exhaustion, task performance and creativity. In Study 2, I examined the main effects of leaders' on-the-job coaching behavior on employees' task and adaptive performance, as well as the moderating effect of task-related expertise and the mediating effects of learning self-efficacy and specific emotions. This experimental study involved two conditions (i.e., guidance versus facilitation coaching) and 209 participants who were coached on how to effectively use a spreadsheet program (Excel 2010) on a personal computer to perform data searching and analysis work in a simulated office environment. Task and adaptive task performance were measured by assessing individual performance on a complex Excel task (task performance) and a complex PowerPoint task (adaptive task performance) after the coaching sessions. Surprisingly, guidance coaching led to better performance than did facilitation coaching on both similar and adaptive tasks. Learning self-efficacy and two types of emotions, (cheerfulness and dejection) mediated these relationships. Furthermore, moderated mediation models with task-related expertise as moderator were also tested. Task-related expertise moderated the mediating effect of learning self-efficacy on the relationship between coaching style and the two outcomes of coaching. The theoretical and practical implications for individual adaptation are then discussed.