Explanatory style, the ways in which people explain both good and bad events (Seligman, 1998), shares theoretical components with teachers' sense of efficacy (Tcshannon-Moran & Woolfolk-Hoy, 2001), which is how capable teachers feel about teaching. According to Bandura (1994), efficacy informs explanatory style, but this assertion does not explain how hard-fought classroom mastery experiences are overcome with little or no efficacy. The three studies presented here suggest that explanatory style mediates teachers' sense of efficacy in predicting burnout in teachers, providing a way to develop efficacy using positive and negative events. Study one provides a conceptual overview of teacher self-efficacy, explanatory style and teacher burnout research and examines the theoretical relationships among these constructs. This study provides the theoretical foundation for studies two and three. In study two, the Educator Attributional Style Questionnaire (EdASQ), based on the more general Attributrional Style Questionnaire (ASQ), was developed to measure teachers' explanatory style. Study two surveyed 350 teachers from three school districts, two of which were used as a cross-validation group for comparison with the other district. The items of EdASQ have high internal reliability and convergent validity, for it correlates with the Rosenberg Self-Esteem Scale (RSES) in a similar fashion as the ASQ. Confirmatory factor analysis supports two distinct measurement models for the EdASQ, one for positive event items and one for negative event items. Study three tests the relationships among explanatory style, teachers' sense of efficacy and teacher burnout. The responses from all the teachers from study two were used for this study. Structural models where explanatory style is a mediator for teachers' sense of efficacy in predicting burnout, as measured by the Oldenburg Burnout Inventory (OLBI) and the number of upcoming professional development hours teachers expect to participate in, were superior to the alternative models. The models suggest that pessimists' explanations of good events mediate their efficacy in predicting burnout while optimists' explanations of bad events mediate their efficacy in predicting burnout. Future research is discussed, including the development of teacher training that capitalizes on explanatory style's role in building efficacy to avoid burnout in teachers.