Background: The clinical effects of prosthesis-patient mismatch (PPM) after aortic valve replacement, with respect to morbidity and survival, remain controversial, particularly in high-risk patient subgroups. Methods: Patients undergoing aortic valve replacement from January 1992 to December 2010 were classified according to effective orifice area index into severe PPM (effective orifice area index < 0.65 cm/m), moderate PPM (effective orifice area index 0.65 to 0.85 cm/m), and absent PPM (effective orifice area index > 0.85 cm/m). Analyses examined major morbidity and total all-cause death. Results: Prosthesis-patient mismatch was classified as severe (92 of 1,060; 8.7%), moderate (440 of 1,060; 41.5%), or absent (528 of 1,060; 49.8%). Moderate and severe PPM were unrelated to in-hospital morbidity or mortality. There were 440 deaths (41.5%) at 5.6 years median follow-up (interquartile range, 2.9 to 9.1). Trend toward poorer survival according to PPM group (χ2 = 5.46; p = 0.07) was attenuated further with covariate adjustment. Sensitivity analyses demonstrated discrete mortality effects for moderate PPM in association with concomitant coronary artery bypass grafting, impaired left ventricular function, and older age (significant hazard ratios range, 1.05 to 1.57). Severe PPM also increased mortality risk in association with older age, concomitant coronary artery bypass grafting, and New York Heart Association Class III or IV (significant hazard ratios range, 1.06 to 2.65). Conclusions: Prosthesis-patient mismatch was not associated with mortality in covariate-adjusted models. However, a discrete mortality risk was attributable to moderate and severe PPM in patients of older age, or those with left ventricular dysfunction, New York Heart Association class III or IV, and concomitant coronary artery bypass grafting.