Social-emotional skills have been shown to be beneficial for many important life outcomes for students. However, previous studies on the topic have suffered from many issues (e.g., consideration of only a small subset of skills, single-informant, and single-cohort design). To address these limitations, this study used a multi-informant (self, teacher, and parent) and multicohort (ages 10-15 from Finland, N = 5,533) perspective to study the association between 15 social-emotional skills and 20 educational (e.g., school grades), social (e.g., relationships with teachers), psychological health (e.g., life satisfaction), and physical health outcomes (e.g., sleep trouble). Results showed that (a) there was a modest level of interrater agreement on social-emotional skills, with the highest agreement between students and parents (mean r = .41); (b) inclusion of multi-informant ratings substantially enhanced the ability of social-emotional skills in predicting outcome variables, with parent- and self-rated skills playing important, unique roles; (c) by modeling skills at the facet level rather than at the domain level, we identified the key skills for different outcomes and found significant variation in facets' predictive utility even within the same domain; and (d) although the older cohort showed lower levels of most social-emotional skills (9/15), there were only minor changes in the interrater agreement and predictive utility on outcomes. Overall, self-control, trust, optimism, and energy were found among the four most important skills for academic and life success. We further identified the unique contribution of each skill for specific outcomes, pointing the way to effective and precise interventions.