The performance of the standard Hancock porcine bioprosthesis (HPB), the Björk-Shiley spherical disc prosthesis (BSP) and the Lillehei-Kaster tilting disc prosthesis (LKP) was compared over a 15-year follow-up in patients who underwent isolated mitral valve replacement (502 HPB, 183 LKP and 119 BSP). The cumulative duration of follow-up is 3,488 patient-years (mean, 8.0 +/- 3.7 years) for patients with a HPB, 1,302 patient-years (mean, 8.4 +/- 4.9 years) for patients with a LKP and 718 patient-years (mean, 7.1 +/- 5.6 years) for those with a BSP. At 15 years a significant difference was found among the 3 groups in freedom from (1) anticoagulant-related haemorrhages, between HPB (83% +/- 7%) and BSP (65% +/- 8%) recipients (P less than 0.05); (2) prosthetic dysfunction, BSP (96% +/- 3%) and LKP (83% +/- 6%) versus HPB (42% +/- 5%) recipients (P less than 0.001); (3) overall reoperations, BSP (94% +/- 3%) and LKP (89% +/- 3%) versus HPB (40% +/- 5%) recipients (P less than 0.001) and (4) prosthetic failure, BSP (88% +/- 4%) and LKP (71% +/- 6%) versus HPB (38% +/- 5%) recipients (P less than 0.001). No difference was found in actuarial survival, freedom from thromboembolic episodes, treatment failure and overall prosthesis-related complications. Our results confirm that the long-term performance of the HPB in the mitral position is adversely affected by the high incidence of tissue degeneration while haemorrhagic complications may limit the advantage of the longer durability of mechanical prostheses.(ABSTRACT TRUNCATED AT 250 WORDS)