Background: Management of atrial fibrillation is frequently challenging, and traditionally catheter or surgical procedures are the main modalities for ablation. Minimally invasive thoracoscopic techniques recently developed endeavor to preserve the advantages of both traditional modalities. We reviewed the efficacy and safety of minimally-invasive MAZE to treat atrial fibrillation in this meta-analysis. Methods: We searched Pubmed and Embase databases from 1 January 1980 to 30 June 2015 for original studies. Two authors evaluated these studies for inclusion independently, and retrieved data for pooled analyses. Results: From the search, 1,098 abstracts were screened with 206 full-articles evaluated for the final inclusion of 31 studies (all observational). Pooled proportion of patients in sinus rhythm at 3 months was 0.85 (0.77-0.93) in 5 studies, at 6 months was 0.88 (0.83-0.94) in 8 studies and at 12 months was 0.84 (0.81-0.97) in 6 studies. Pooled proportion of patients off anti-arrhythmic drugs at 3 months was 0.52 (0.01-1.03) in 3 studies, at 6 months as 0.73 (0.64-0.83) in 9 studies and at 12 months 0.67 (0.58-0.77) in 11 studies. Adverse event rates of mortality (0-2%), stroke/TIA (0-3%) and bleeding (0-15%) in studies reported were low. Conclusion: Efficacy and safety of minimally-invasive MAZE is high and similar to catheter and surgical ablation. Ongoing development and larger and randomised studies of this procedure would facilitate more precise comparisons with other treatment modalities and guide clinical practice.