This thesis examines the effectiveness of interventions to reduce school- and cyberbullying in two meta-analytical reviews. Bullying is defined as repetitive and intentional behaviours that involve children and adolescents between whom there is a distinct power imbalance. Cyberbullying is the manifestation of these behaviours through information and communications technologies (e.g., mobile phones, social media, online environments). Evaluations of school-based anti-bullying programmes designed to reduce offline and online bullying were reviewed. The school-bullying meta-analysis included 100 evaluations of anti-bullying programmes and found that overall, interventions to reduce and prevent bullying were effective. Included programmes significantly reduced school-bullying perpetration by approximately 19-20% and school-bullying victimisation by approximately 15-16%. Several intervention components (e.g., whole-school approach, anti-bullying policies, classroom rules, and informal peer involvement) were associated with greater reductions in school-bullying perpetration. Interventions that included informal peer involvement (e.g., group discussion, role play, class exercises) and information for parents (e.g., leaflets or letters) were associated with greater reductions in school-bullying victimisation. The cyberbullying meta-analysis found that interventions significantly reduced cyberbullying perpetration by approximately 10-15% and cyberbullying victimisation by approximately 14%. This thesis also provides support for a new method of assigning weight to primary studies in a meta-analysis, the Multiplicative Variance Adjustment model, and explores many avenues for future research on 'what works' in bullying intervention and prevention.