Burnout is prevalent among mental health providers and is associated with significant employee, consumer, and organizational costs. Over the past 35 years, numerous intervention studies have been conducted but have yet to be reviewed and synthesized using a quantitative approach. To fill this gap, we performed a meta-analysis on the effectiveness of burnout interventions for mental health workers. We completed a systematic literature search of burnout intervention studies that spanned more than 3 decades (1980 to 2015). Each eligible study was independently coded by 2 researchers, and data were analyzed using a random-effects model with effect sizes based on the Hedges’ g statistic. We computed an overall intervention effect size and performed moderator analyses. Twenty-seven unique samples were included in the meta-analysis, representing 1,894 mental health workers. Interventions had a small but positive effect on provider burnout (Hedges’ g = .13, p = .006). Moderator analyses suggested that person-directed interventions were more effective than organization-directed interventions at reducing emotional exhaustion (Qbetween = 6.70, p = .010) and that job training/education was the most effective organizational intervention subtype (Qbetween = 12.50, p < .001). Lower baseline burnout levels were associated with smaller intervention effects and accounted for a significant proportion of effect size variability. The field has made limited progress in ameliorating mental health provider burnout. Based on our findings, we suggest that researchers implement a wider breadth of interventions that are tailored to address unique organizational and staff needs and that incorporate longer follow-up periods.