Limited data are available on mitral valve replacement in the elderly patient. Therefore we report our 10-year experience including predictors of perioperative mortality and subsequent long-term cardiac mortality in elderly patients with mitral valve replacement compared to younger patients. Of the 126 consecutive patients with mitral valve replacement, 26 were older (77 +/- 4, group 1) and 100 were younger (62 +/- 9, group 2) than 70 years. Bioprostheses were used more frequently in patients in group 1 (65% vs 7%, p less than 0.0001). Of the 21 clinical, ECG, hemodynamic, and angiographic variables studied, patients in group 1 had higher pulmonary artery systolic pressure (57 +/- 15 vs 48 +/- 19, p less than 0.05), fascicular block on ECGs (70% vs 33%, p less than 0.005), and greater pump time on cardiopulmonary bypass (160 +/- 75 vs 120 +/- 50 minutes, p less than 0.025). A trend toward a higher perioperative mortality rate was also seen in group 1 (27% vs 12%, p = 0.058). Predictors of perioperative mortality by multivariate analysis were the presence of aortic calcification and prolonged pump time on cardiopulmonary bypass in group 1 and coronary artery disease, female sex, elevated mean pulmonary artery pressure, and postoperative complete atrioventricular block in group 2. During a mean 4-year follow-up period, cardiac mortality and total mortality rates were 42% and 54%, respectively, for group 1 compared to 24% and 35%, respectively, for group 2.(ABSTRACT TRUNCATED AT 250 WORDS)