A questionnaire was sent to all nurses of a university hospital. Usable returns were received from 625 nurses, giving a response rate of 39.2%. Data were treated using correlational analyses and multiple regression. The latter modelled stressors and leadership as predictors of nurses' reported emotional exhaustion. Work stressors were assessed using the Nursing Stress Scale (NSS) which comprises 34 items divided into three subscales (referring to stress from the physical, psychological, and social environment), and the role ambiguity (three items) and conflict (three items) scales. Leadership was measured with the Multifactor Leadership Questionnaire. In regression analyses, work stressors as a whole were found to explain 22% of the variance in emotional exhaustion whereas leadership dimensions explained 9% of the variance in that outcome measure. Stress emanating from the physical and social environment, role ambiguity, and active management-by-exception leadership were significantly associated with increased levels of emotional exhaustion. Transformational and contingent reward leadership did not influence emotional exhaustion. A limitation of this study is that it considered only the emotional exhaustion dimension of burnout. Also, as data were cross-sectional in nature, conclusions regarding the direction of causality among variables cannot be drawn. This study provided, for the first time, a test of the influence of leadership on burnout among nurses, taking into account the role of work stressors. Future research is needed to examine if the effects reported herein can be replicated using the two other dimensions of burnout (depersonalization and reduced personal accomplishment).