Background: The purpose of this study was to determine the impact of prosthesis-patient mismatch on long-term survival after mitral valve replacement. Methods: From 1992 to 2008, 765 patients underwent bioprosthetic (325; 42%) or mechanical (440; 58%) mitral valve replacement, including 370 (48%) patients older than 65 years of age. Prosthesispatient mismatch was defined as severe (prosthetic effective orifice area to body surface area ratio <0.9 cm2/m2), moderate (0.9 to 1.2 cm 2/m2), or absent (>1.2 cm2/m2). Results: Multivariate analysis identified nine risk factors for late death including advanced age, earlier operative year, chronic renal insufficiency, peripheral vascular disease, congestive heart failure, nonrheumatic origin, concomitant coronary artery bypass grafting, lower body surface area, and more severe prosthesispatient mismatch (lower effective orifice area to body surface area ratio; p < 0.05). For bioprosthetic recipients older than 65 years of age, survival at 5 and 10 years was 30% ± 7% and 0% ± 0% with severe mismatch compared with 43% ± 4% and 21% ± 5% for absent or moderate mismatch, respectively (p = 0.05). For mechanical recipients younger than 65 years of age, survival at 5 and 10 years was 77% ± 4% and 62% ± 6% with moderate or severe mismatch compared with 82% ± 3% and 66% ± 4%, respectively, without mismatch (p = 0.08). Conclusions: Severe mismatch adversely affected long-term survival for older patients receiving bioprosthetic valves. With mechanical valves, there was a trend toward impaired survival when mismatch was moderate or severe in younger patients. Thus, selection of an appropriate mitral prosthesis warrants careful consideration of age and valve type.