Over the last few years, the transformational, transactional and laissez-faire leadership theory has attracted the attention of many researchers (Judge and Piccolo, 2004; Lowe and Gardner, 2000, Yammarino et al., 2005). Furthermore, numerous studies (Bycio, Hackett and Allen, 1995; Dumdum, Lowe and Avolio, 2002; Rafferty and Griffin, 2004; Yammarino, Spangler and Dubinsky, 1998) have shown that these three forms of leadership have a significant impact on several employee attitudes and behaviours, including affective organizational commitment, which is characterized by an individual's emotional attachment to his company (Meyer and Allen, 1997). Nevertheless, there are still very few researchers taking an interest in the processes through which leaders can produce such effects (Bass and Riggio, 2006; Bono and Judge, 2003; Yukl, 2006). Only a few authors have tackled this problem, particularly by identifying empowerment (Avolio et al., 2004) and fairness (Pillai, Schriesheim and Williams, 1999) as mechanisms explaining the relationship between transformational leadership and employees' affective commitment. We intend to build on these recent results in improving the understanding of this dynamic. Based on the social exchange theory (Blau, 1964), this research is aimed at exploring how supervisors can strengthen their employees' affective commitment. More specifically, we will assess the extent to which support and trust can account for the influence of each of the dimensions associated with transformational, transactional and laissez-faire leadership has on commitment. These dimensions are: charisma, intellectual stimulation, individualized consideration, contingent reward, active management by exception, and passive avoidance. One of the primary contributions of this article resides in the fact that currently, there are still very few empirical research projects that have focussed on the effects of these dimensions on other variables, and that such research has only concentrated on the influence of global forms of leadership, and almost exclusively on the transformational form. However, recent works have indeed highlighted the importance of using specific dimensions rather than these three major forms of leadership since the later provide an imperfect and oversimplified image of all the leaders' behaviours and potential (Antonakis, Avolio and Sivasubramaniam, 2003). In addition to filling a gap at this level, the second significant contribution of this project lies in the fact that, to our knowledge, no study has as yet validated the role of support as an intermediate mechanism lying between leadership and affective commitment. In concrete terms, we are first proposing that the charisma, intellectual stimulation, individualized consideration and contingent reward dimensions are positively related to employees' perceived supervisor support and trust in their supervisor, whereas the active management by exception and passive avoidance dimensions are negatively related to those same two variables. Secondly, we are proposing that employees' perceived organizational support and trust in their organization act as mediating variables between perceived supervisor support and trust in their supervisor and affective commitment, respectively. This research, which was conducted in the spring of 2004 among all the employees (excluding physicians) of a Quebec hospital center, allowed the collection of 568 questionnaires, representing a 46% response rate; 80% of the subjects in the sample were women, whose average age was 43, and who had been working in the organization for 12 years, on average. To test our hypotheses, confirmatory factorial analyses and structural equations were conducted, while controlling for gender and employment status (full time vs. part time). The results of the study indicate that only the charisma dimension seems to lead employees to trust their supervisor. This observation is particularly interesting for researchers with an interest in interpersonal trust since other studies have arrived at similar conclusions (Gillespie and Mann, 2004). However, our results indicate that charisma and contingent reward are positively related to perceived supervisor support whereas active management by exception is negatively associated to it. This observation is particularly important in that it provides responses to the theoretical arguments which were paving the way for such a possibility (Yammarino and Bass, 1990; Jung and Avolio, 2000). Lastly, this research shows that the constructs of perceived supervisor support and perceived organizational support constitute an important explanatory mechanism in the relationship between leadership and affective commitment. Trust in the organization also contributes to explaining the dynamic existing between leadership and commitment; however, it is not significantly influenced by trust in the supervisor. Our results open up several avenues of further research. Although our analyses have identified support and trust as intermediate mechanisms between leadership and commitment, other mediators could still account for this relationship. In addition to exploring this avenue, future research could simultaneously analyze the intermediate variables that have been identified to date in the literature (e.g., support, trust, fairness, empowerment) with respect to commitment, but also to other consequences which are often related to transformational and transactional leadership (e.g., satisfaction, mobilization, performance), in order to better understand their relative importance. Finally, this study has certain limitations, including the difficulty in generalizing results (sample composed of a single organization), the possible inflation of the strength of certain relationships (common variance bias) and the impossibility of inferring the causality of the observed relationships (cross-sectional design).