Previous reviews have synthesized the impacts of universal school-based social emotional learning (SEL) programs. However, they have yet to attempt a meta-analytic approach with rigorous inclusion criteria to identify the key SEL components and explore what make these programs work. This study aims to fill that gap by examining the impacts of SEL programs and exploring the moderating effects of methodological characteristics, implementation features, and program components on SEL effectiveness. The final sample consisted of 12 high-quality SEL programs, 59 studies, and 83,233 participants, with an overall effect size of 0.15. Meta-regression results indicated that these SEL programs could significantly improve youth social emotional skills, reinforce affect and attitudes, promote academic performance, increase prosocial behaviors, and reduce antisocial behaviors. Training teachers' social emotional skills and reducing cognitive elements in SEL curricula were found to be effective components of SEL programs, whereas pedagogical activities, climate support, and family engagement were not. Large-scale studies of SEL programs tended to generate smaller effect sizes, and those with low program dosages were found to be less effective than those approaching the recommended dosage. Policy and practical implications on how to scale SEL programs are discussed.