Hancock porcine bioprostheses have been implanted in the mitral position at the National Institutes of Health since July, 1970. Eight models (330, 330A, 330B, 330C, 332, 340, 341, and 342) were used during a 54-month period ending December, 1974, and 100 consecutive surviving patients were evaluated for subsequent bioprosthetic valve failure and prosthesis-related complications by annual clinic examinations and serial hemodynamic studies. Actuarial patient survival was 76 ± 4%, 51 ± 5%, and 30 ± 6% after 5, 10, and 15 years, respectively. Intrinsic valve failure, defined as structural degeneration of bioprosthetic tissue or stent geometry alteration or both, in the absence of prior infection, occurred in 23 patients. The linear occurrence rate of bioprosthetic valve failure was 0.2%, 5%, and 15% per patient-year, and it affected 1 patient, 14 patients, and 8 patients at sequential 5-year milestones. The actuarial freedom from valve failure was 99 ± 1%, 75 ± 6%, 58 ± 8%, and 40 ± 12% after 5, 10, 12, and 14 years, respectively. The valve durability of early Hancock bioprostheses (models 330 through 341; N = 39) was not appreciably different from that of the current model 342 valves (N = 61). However, an increased incidence of intrinsic valve failure was observed for the first polypropylene-stented valve type (model 330) compared with the currently available model 342 valve (8/16, 50%, versus 12/61, 20%; p = 0.034). The yearly occurrence rate of prosthesis-related complications remained constant, but the rate of intrinsic valve failure increased in a progressive, nonlinear fashion. The high intrinsic failure rate of the Hancock porcine bioprosthesis after 10 to 12 years has moderated our initial enthusiasm for this valve in the mitral position, and has resulted in more frequent implantations of mechanical valve substitutes at this institution.