A retrospective study was conducted on 124 patients who underwent re-replacement of previously implanted prosthetic heart valves for structural valve failure, prosthetic valve endocarditis, periprosthetic leak, a thrombosed valve, hemolysis, or prophylactic removal. In total, 85% of the explanted valves were bioprostheses, and 70% of the newly implanted valves were mechanical valves. The overall operative mortality rate was 8.1%, being 3.2% of 95 single valve recipients and 25.0% of 28 double valve recipients (P < 0.001). The overall mortality rate dropped from 13.6% of 66 patients before 1988, to 1.7 % of 58 patients encountered in the last 3 years (P < 0.02). Since 1988, a third of the patients have undergone reoperation without homologous blood transfusion. A univariate analysis revealed eight operative risk factors, namely: higher values of preoperative blood urea nitrogen or total bilirubin, double valve replacement at the redo operation, NYHA class IV, urgency of reoperation, a duration of implantation of less than 3 months, reoperation in the earlier period of this study, and reexploration for bleeding or cardiac tamponade after re-replacement surgery. A multivariate statistcial analysis demonstrated that preoperative blood urea nitrogen, urgency of reoperation, double valve replacement, and the duration of implantation were independent risk factors. Thus, we recommend that surgery be performed early, before the occurrence of other organ failure induced by congestive heart failure due to any form of valve dysfunction.