The purpose of this study was to justify and predict the self-regulation learning based on Bandura social cognitive model. The research sample consisted of 384 1st and 2nd grade senior high school students of Urmia (204 male and 180 female students) who were selected based on multi-stage random sampling method. Kerjcie and Morgan chart was used to determine the sample size. To evaluate the variables, participants completed the tests of positive and negative affect scales, perception of school climate, self-efficacy questionnaire, outcome expectations, achievement goals and self-regulation learning strategies questionnaire.Data was analyzed by structural equation modeling. The results indicated that the self-regulation learning model is a valid and reliable model for this community. The variables (i.e., school climate, outcome expectations and achievement goals) could predict 65% of the self-regulation learning. Path coefficient analysis displayed that academic self-efficiency has the highest impact on self-regulation learning. Accordingly, it can be concluded that the self-regulation learning model explicate the social cognitive determinants of students’ self-regulation learning, and we can use it to promote self-regulation learning, learning quality and academic achievement of students at school.