All patients who had a mitral or aortic Hancock valve replacement between June, 1974, and June, 1979, were reviewed. A total of 734 bioprostheses were implanted in 632 patients: 291 had mitral (MVR), 239 had aortic (AVR), and 102 had both mitral and aortic valve replacement (MVR + AVR). In 228 patients, an associated surgical procedure was necessary. It involved conservative valve operation in 205 of them. The hospital mortality was 9.6% (28) for MVR (11.5% with associated operation), 4.6% (11) for AVR (8.7% with associated operation), and 13.7% (14) for MVR + AVR (13.0% with associated operation). The follow-up period was between 1 and 6 years with a total follow-up of 934.6, 714.6, and 288.3 patient-years for MVR, AVR and MVR + AVR, respectively. The late mortality was 0.96% (9), 1.53% (11), and 2.08% (6) per patient-year for MVR, AVR, and MVR + AVR, respectively. The thromboembolic rate was 1.49%, 0.14%, and 2.08% per patient-year for MVR, AVR, and MVR + AVR, respectively. There were twelve valve failure (six wer due to rupture; four, thrombosis; one, insufficiency because of intrinsic failure; and one, stenosis without evident cause at reoperation). This represents a failure rate of 0.53%, 0.13%, and 2.08% per patient-year for MVR, AVR, and MVR + AVR, respectively. These results encourage us to continue our routine use of the glutaraldehyde xenografts as the safest valve substitute at present.