A total of 1,239 patients had cardiac valve replacement using 1,514 substitute valves at Seoul National University Hospital from 1968 to 1986. Of the total substitute vales, 84.9% were the glutaraldehyde-treated xenograft valves. Six hundred ninety-four patients who had 820 bioprosthetic tissue valves were studied for their clinical characteristics. They were a total and consecutive cases to the end of the study. Four hundred sixty-four patients had the lonescu-Shiley pericardial valves: MVR 291, AVR 66 and MVR+AVR 107; 163 had the Hancock porcine valves; 46 had the Angell-Shiley porcine valves; and 21 had the Carpentier-Edwards porcine valves. Five hundred forty patients underwent single valve replacement: MVR 460, AVR 76 and TVR 4; 154 had multiple valve replacement: MVR+AVR 141, MVR+TVR 12 and one triple valve replacement. Additional surgery was necessary in 22.3% of the cases. Operative mortality rate within 30 days of surgery was 6.77% for the total patients: 5.2% and 4.2% with MVR, 13.6% and 12.5% with AVR, and 7.5% and 7.4% with MVR+AVR using the lonescu and the Hancock valves respectively. A linealized annual late mortality rate was 2.56%/patient-year. Six hundred forty-three operative survivors were followed up for a total of 1482.7 patient-years [a mean 27.7 months], and the follow-up rate was 67.7%. The Idealized complication rates were: 2.02% emboli/patient-year, 0.94% bleeding/patient-year, 1.21% endocarditis/patient-year, and 3.84% overall valve failure/patient-year. A linealized rate of primary tissue failure was 0.87%/patient-year. Actuarial survival rates including the operative mortality were: 87.8*2.6%, 82.3*4.9% and 82.2*4.7% with MVR, AVR and MVR+AVR using the lonescu valves at 4 years after surgery respectively; and they were 88.0*4.1% with MVR at 8 years, 82.3*4.9% with AVR at 4 years and 84.9*7.0% with MVR+AVR at 6 years after surgery using the Hancock valves respectively. Probabilities of freedom from thromboembolism were 89.8*6.3% with MVR using the lonescu valves at postoperative 5 years and 89.2*3.8% with MVR using the Hancock valves at postoperative 7 years, and 93.3*3.9% with AVR using the lonescu valves at postoperative 5 years. None had embolic complication after AVR using the Hancock valves. Probabilities of freedom from valve failure [according to the Stanford criteria] were 81.0*7.1% with MVR using the lonescu valves at postoperative 4 years and 57.4*12.5% with MVR using the Hancock valves at postoperative 9 years. These clinical results prove the excellent antithrombogenicity of the glutaraldehyde-treated xenograft substitute valves and confirm the previously speculated rate of tissue failure. At the present situation, it may be concluded that there is a room for the further development of more durable bioprosthetic valves.