Background: We tried to identify clinical and echocardiographic predictors of clinical outcome in severe, asymptomatic rheumatic mitral regurgitation (MR). Methods: From 1997 to 2006, we prospectively enrolled a total of consecutive 180 patients (33 men, age;41짹15 yrs) with severe rheumatic MR, and the exclusion criteria were defined as the presence of exertional dyspnea, ejection fraction <0.60, significant mitral stenosis or aortic valve disease. Severe rheumatic MR was defined as a restricted motion and thickening of mitral valve (MV) with the radius of proximal isovelocity surface area of MR > 8 mm documented by echocardiography. Early surgery was performed on 57 patients within 6 month of enrollment, and conventional strategy was chosen for 123 patients. The endpoint of the study was defined as cardiac death during follow-up. Results: During the median follow-up of 2256 days, there were 9 cardiac deaths and one non-cardiac death. Age (hazard ratio 1.11, p < 0.01) and serum creatinine level (hazard ratio 2.68, p = 0.02), were independent variables associated with cardiac mortality. It was not significant whether the patient underwent early surgery or conventional treatment in cardiac mortality (p = 0.23). In 123 patients of the conventional treatment group, 41 patients underwent late MV surgery and there was no operative mortality. LA dimension (hazard ratio 1.04, p=0.03) was the independent variable associated with development of late OP indication. In total of 98 patients who underwent MV surgery either early or late, age (odds ratio 0.95, p = 0.03) and echo score (odds ratio 0.21, p < 0.01) were unfavorable predictors for successful mitral repair. Conclusion: In a selected group of asymptomatic severe rheumatic MR, age and serum creatinine level were predictive factors for cardiac death. Successful MV repair was influenced by age, echo score and not by the timing of surgery.