Between July 1984 and December 1991, 541 Carpentier-Edwards pericardial prostheses were implanted in 536 patients (420 aortic and 121 mitral prostheses). The operative mortality was 2.9% for aortic and 3.3% for mitral valve replacement. All but 8 patients were followed up for an average of 3.9 years, making a total follow-up of 1930 patient-years. There were 61 deaths during follow-up (41 AVR, 20 MVR). The actuarial 8 year survival was 80 +/- 6% for AVR and 74 +/- 10% for MVR. There were 15 deaths related to the prostheses. The probability of absence of valve-related mortality was 97 +/- 3% for AVR and 92 +/- 6% for MVR (NS). The valve-related complications included 15 thromboembolic events (AVR: 12; MVR: 3), 10 endocarditis (AVR: 7; MVR: 3), 8 anticoagulant-related haemorrhages (AVR: 4; MVR: 4), 4 degeneration (AVR: 2; MVR: 2). There were no cases of cusp tear. The probability of absence of thromboembolic events at 8 years was 96 +/- 3% for AVR and 96 +/- 4% for MVR (NS), of absence of endocarditis 97 +/- 2% for AVR and 97 +/- 3% for MVR (NS) and of absence of valve degeneration 99 +/- 1% for AVR and 99 +/- 2% for MVR (NS). These results show that the position of the bioprosthesis had no influence on valve-related morbidity and that pericardium is a satisfactory tissue for the manufacture of mitral bioprostheses.