ObjectiveWhether mitral valve repair is superior to replacement in the population with rheumatic heart disease has been debated. This study aims to compare outcomes of repair with replacement by the propensity score method. MethodsThis observational, prospective study enrolled patients with rheumatic heart disease who underwent mitral valve repair and replacement from January 2011 to April 2019. The propensity score method was used to select 2 groups with similar baseline characteristics. Baseline, clinical, and follow-up data were collected. Clinical outcomes included death from any cause, reoperation, and valve-related complications. ResultsThe overall population before matching (N = 1644) included 612 patients who underwent repair and 1032 patients who underwent replacement. The propensity score analysis generated matches for 1058 patients (529 pairs). The median follow-up time was 4.12 years. Early mortality and death from any cause during follow-up were significantly lower in the repair group compared with the replacement group (hazard ratio, 0.19; 95% confidence interval [CI], 0.05-0.64; P = .003; hazard ratio, 0.38; 95% CI, 0.19-0.74; P = .003, respectively). Patients in the repair group had a lower risk of valve-related complications compared with patients in the replacement group (subhazard ratio, 0.44; 95% CI, 0.21-0.90; P = .025). In terms of reoperation, no significant difference was observed between the repair and replacement groups (subhazard ratio, 2.54; 95% CI, 0.89-7.22; P = .081). ConclusionsThe results suggest that rheumatic mitral valve repair in select patients is superior to mitral valve replacement with regard to lower mortality and fewer valve-related complications; meanwhile, it has a comparable risk of reoperation compared with replacement.