Objectives Self-compassion is a healthy way of relating to one’s self motivated by a desire to help rather than harm. Novel self-compassion-based interventions have targeted diverse populations and outcomes. This meta-analysis identified randomized controlled trials of self-compassion interventions and measured their effects on psychosocial outcomes. Methods This meta-analysis included a systematic search of six databases and hand-searches of the included study’s reference lists. Twenty-seven randomized controlled trials that examined validated psychosocial measures for self-compassion-based interventions met inclusion criteria. Baseline, post and follow-up data was extracted for the intervention and control groups, and study quality was assessed using the PRISMA checklist. Results Self-compassion interventions led to a significant improvement across 11 diverse psychosocial outcomes compared with controls. Notably, the aggregate effect size Hedge’s g was large for measures of eating behavior ( g  = 1.76) and rumination ( g  = 1.37). Effects were moderate for self-compassion ( g  = 0.75), stress ( g  = 0.67), depression ( g  = 0.66), mindfulness ( g  = 0.62), self-criticism ( g  = 0.56), and anxiety ( g  = 0.57) outcomes. Further moderation analyses found that the improvements in depression symptoms continued to increase at follow-up, and self-compassion gains were maintained. Results differed across population type and were stronger for the group over individual delivery methods. Intervention type was too diverse to analyze specific categories, and publication bias may be present. Conclusions This review supports the efficacy of self-compassion-based interventions across a range of outcomes and diverse populations. Future research should consider the mechanisms of change.