Optimistic attributional style has been shown to be reliably associated with high well-being and low depression and (Hu et al., "Journal of Social and Clinical Psychology," 34(4): 304-321, 2015). Via both a meta-analysis and two new studies, we examine the relationship between optimistic attributional style for explaining negative and positive events, and academic performance. In the meta-analysis, dispositions to make stable and global attributions for positive events were more strongly related to academic achievement (d = 0.21, k = 30, N = 6351) than dispositions to make unstable and local attributions regarding negative events (d = 0.11, k = 66, N = 11,023). Academic level (primary school vs secondary school vs university) and type of test (general vs achievement-specific) were shown to moderate the associations. The two new studies were designed to address the remaining questions. In both studies, optimistic attributional style for positive events most reliably predicted student academic achievement, including boosted achievement over time. Possible explanations for the moderator effects are discussed, and recommendations for future research as well as practical recommendations are provided.