Background Advances in tissue prosthetic valve design and manufacturing have stimulated renewed interest in the use of biological valves in younger patients. This approach, however, risks reoperation. We therefore reviewed our recent experience with repeat mitral valve replacement to better define its contemporary risks. Methods Using a computerized database, we identified and compared 106 patients undergoing repeat mitral valve replacement with 562 control patients undergoing primary mitral valve replacement between January 1993 and December 2000 at our institution. Results There were no significant differences between repeat and primary surgery groups with respect to age (mean 66 ± 12 vs 64 ± 13 years), gender distribution (women 65% vs 64%), preoperative functional class, ejection fraction, or active endocarditis (6.6% vs 3.4%). The indication for reoperation in the repeat group was structural dysfunction in 49 patients (46%), paravalvular leak in 21 patients (20%), nonstructural dysfunction in 11 patients (10%), and progression of other native valve disease in 8 patients (8%). Prior prostheses were mechanical in 46 patients (43%). Mean time to reoperation was 11.5 ± 7.1 years. There were 5 deaths out of 106 patients in the repeat group (4.7%) and there were 23 deaths out of 562 patients in the control group (4.1%) (p = NS). Multivariate analysis identified prior myocardial infarction (p = 0.014, odds ratio 2.9) and nonelective surgical status (p = 0.004, odds ratio 2.3) as significant predictors of operative mortality. Conclusions The risk of repeat mitral valve replacement was low suggesting that there should be less reluctance to recommend patients choose a bioprosthesis over a mechanical prosthesis. Given the expected durability of current designs, bioprosthetic use may be explored in younger patients without subjecting those individuals to excessive risk.