Background We compared 30-day and 1-year survival among high-risk mitral regurgitation (MR) patients treated with the MitraClip (Abbott Vascular, Abbott Park, IL) with matched non-surgically treated patients from the Duke Echocardiography Laboratory Database (DELD). Methods and Results High-risk patients with 3+/4+ MR managed non-surgically between years 2000 and 2010 in the longitudinal DELD were matched to high-risk MitraClip patients. Patient matching was performed using the method of nearest available Mahalanobis distance metric within calipers defined by the propensity score. Kaplan-Meier estimates and stratified Cox proportional hazards models were used to compare survival at 30 days and 1 year. Among 953 high-risk DELD patients available for matching, 30-day and 1-year mortality were 6.5% and 26.2%. Close matches were obtained for 239 of the 351 MitraClip patients. The 30-day mortality in MitraClip patients was lower (4.2%) when compared with matched DELD patients (7.2%). The 1-year relative risk of mortality of the MitraClip compared with non-surgical treatment was 0.64 (95% CI 0.45-0.91; log-rank P = .013). These results in favor of the MitraClip remained significant upon further adjustment for baseline differences between groups ( P = .043). Conclusions This matched comparison of severe MR patients at high surgical risk supports the safety of the MitraClip relative to medical therapy at 30 days and a survival benefit at 1 year.