Background: Academic self-efficacy, when operationalized as mastery over domain-specific knowledge, has been found to be a predictor of academic achievement and emotions. Although academic emotions are also a predictor of academic achievement, there is limited evidence for reciprocal relations with academic achievement. Aims: To examine whether academic self-efficacy, when operationalized as confidence in study-related skills and behaviours, is also a predictor of academic achievement and emotions and to test reciprocal relations between academic emotions and achievement. Sample: Two hundred and six first-year undergraduate students. Methods: Academic self-efficacy was measured at the beginning of the first semester and learning-related emotions (LREs) at the beginning of the second semester. Academic performance was aggregated across assessments in semester one and semester two. Results: Self-efficacy in study-related skills and behaviours predicted: (1) better semester one academic performance and (2) more pleasant and fewer unpleasant LREs at the beginning of the second semester directly and (3) indirectly through semester one academic performance. Reciprocal relations between academic performance and emotions were supported, but only for pleasant emotions. Conclusions: Self-efficacy in study-related skills was the critical academic self-efficacy variable in this study. It may play an important role in maintaining challenge appraisals to maintain pleasant emotions and better academic performance. Accordingly, practitioners in higher education may wish to consider the value of assessing and developing students' self-efficacy in relation to their independent study skills.