This study suggests that stressors can be productive for self-efficacy and that the influence of stressors on self-efficacy is nonlinear. Analyses were conducted with ordinary least squares regression on a dataset covering responses from 311 deans in Swedish secondary schools. Results support the hypothesized U-shape relationship between role conflict and self-efficacy and the inverted U-shape relationship between role ambiguity and self-efficacy. Thus, findings offer evidence for nonlinear effects of stressors on the level of incumbents' self-efficacy. This research has implications for further research focused on the association between role stressors and self-efficacy.