OBJECTIVE: Mindfulness-based interventions (MBIs) are being increasingly used as interventions for eating disorders including binge eating. This systematic review and meta-analysis aimed to assess two decades of research on the efficacy of MBIs in reducing binge eating severity. METHODS: We searched PubMed, Scopus and Cochrane Library for trials assessing the use of MBIs to treat binge eating severity in both clinical and non-clinical samples. The systematic review and meta-analysis was pre-registered at PROSPERO (CRD42020182395). RESULTS: Twenty studies involving 21 samples (11 RCT and 10 uncontrolled samples) met inclusion criteria. Random effects meta-analyses on the 11 RCT samples (n = 618: MBIs n = 335, controls n = 283) showed that MBIs significantly reduced binge eating severity (g = -0.39, 95% CI -0.68, -0.11) at end of trial, but was not maintained at follow-up (g = -0.06, 95% CI, -0.31, 0.20, k = 5). No evidence of publication bias was detected. On the Cochrane Risk of Bias Tool 2, trials were rarely rated at high risk of bias and drop-out rates did not differ between MBIs and control groups. MBIs also significantly reduced depression, and improved both emotion regulation and mindfulness ability. CONCLUSION: MBIs reduce binge eating severity at the end of trials. Benefits were not maintained at follow-up; however, only five studies were assessed. Future well-powered trials should focus on assessing diversity better, including more men and people from ethnic minority backgrounds.