Background <p>Bullying represents a critical global health concern among children and adolescents, emphasizing an urgent need for effective prevention interventions. Current meta-analyses frequently focus on single prevention interventions or outcomes, limiting the generalizability of findings for cross-intervention comparisons. We aimed to perform an umbrella review to provide an overview of the effectiveness of bullying prevention interventions.</p> Methods <p>Potential literature was searched from PubMed, Web of Science, Embase, and EBSCO from their respective inception to October 22, 2024. Three reviewers independently selected meta-analysis of bullying prevention intervention that reported any type of prevention intervention and assessed outcomes in children and adolescents. Following PRISMA guidelines, three reviewers independently extracted data and evaluated study quality using AMSTAR-2. All effect estimates were converted to Cohen’s d for cross-domain comparisons.</p> Results <p>4740 studies were identified with 30 meta-analyses included. Studies reporting statistical significance all demonstrated beneficial effects of bullying prevention interventions. Included meta-analyses evaluated 17 distinct prevention interventions in alleviating bullying via assessment on three main outcomes with 49 distinct sub-outcomes, i.e., behavior change (36 sub-outcomes, the most frequent target), mental health (9 sub-outcomes), and attitude change (4 sub-outcomes). 30 studies were categorized into five categories based on the leading role of prevention interventions: family-based (<i>n</i> = 2), school-based (<i>n</i> = 10), community-based (<i>n</i> = 2), healthcare-based (<i>n</i> = 5), and collaborative (<i>n</i> = 11). Cohen’s d values ranged from −0.99 to 1.14, with the music-based intervention exhibiting the largest observed point estimate (Cohen’s d = −0.99, 95% confidence interval [CI]: −1.42 to −0.56), which was also the largest among interventions targeting behavior change. Martial arts presented largest observed point estimate among collaborative prevention interventions, while digital health intervention, anti-aggression program, and parent-related intervention showed the largest observed point estimates among community-based, school-based, and family-based prevention interventions, respectively. For mental health, sports intervention exhibited the largest observed point estimate. And for attitude change, anti-bullying intervention presented the largest estimate, consistent with results from RCT-only subgroup.</p> Conclusions <p>Bullying prevention interventions may be associated with improvements in positive behavior change, attitude change, and mental health, with substantial variations in effectiveness. These findings regarding comparative efficacy may inform the strategic selection of preventive interventions in future evaluation and implementation efforts.</p> PROSPERO registration number <p>CRD42024601712</p>