This meta-analysis of eating disorder prevention programs found that intervention effects ranged from an absence of any effects to reductions in current and future eating pathology. Certain effects persisted as long as 2 years and were superior to minimal-intervention control conditions. Larger effects occurred for selected (vs. universal), interactive (vs. didactic), and multisession (vs. single session) programs; for programs offered solely to females and to participants over age 15; for programs without psychoeducational content; and for trials that used validated measures. The results identify promising prevention programs and delineate sample, format, and design features that are associated with larger effects, but they suggest the need for improved methodological rigor and statistical modeling of trials and enhanced theoretical rationale for interventions.