Echocardiographic and Doppler studies were performed in 134 patients with a Hancock bioprosthesis in the mitral valve position during a follow-up period of 1 to 216 months. Among the xenografts, 57% were clinically normal and 43% had severe dysfunction. Among the normal bioprostheses, 35% had echocardiographically thickened mitral cusps (> or = 3 mm) with normal hemodynamic function; by setting the lower 95% confidence limit of valve area at 1.7 cm2 these patients had a significantly (p < 0.01) smaller valve area than that of normal control subjects. Evaluation of all thickened normal mitral valves showed the highest incidence of thickening at 9 years after implantation. Valve replacement surgery was subsequently performed in 33 patients with dysfunctioning bioprosthetic and echocardiographic diagnosis was confirmed in 91% of explanted valves (bioprosthetic stenosis 21%, incompetence 46%, and combined stenosis and regurgitation 33%). In 2 valves that were found to be stenotic on echocardiographic examination, a calcium-related commissural tear was also observed at reoperation, and in another, a paravalvular leak was found. Dystrophic calcification, isolated (64%) or occasionally associated with fibrous tissue overgrowth (21%), was the main cause of failure. Pannus was present in prostheses with longer satisfactory function (168 +/- 31 vs 124 +/- 21 months; p < 0.001). Long-term performance was evaluated by the Kaplan-Meier method for up to 18 years of follow-up. Freedom from structural valvular disfunction after mitral replacement was 89% at 6 years, 77% at 8 years, 56% at 10 years, 31% at 12 years, 16% at 15 years, and 15% at 18 years.