Long-term results of mitral valve replacement using glutaraldehyde-treated porcine bioprostheses were evaluated. The subjects were 77 patients with the Hancock valve (Hancock group) and 60 with the Liotta valve (Liotta group) who survived operation. The maximum follow-up was 14 years and 8 years, and the cumulative follow-up was 631 patient-year (p-y) and 301 p-y in the Hancock and Liotta groups. The actuarial survival rate at 8 years was 80.1 +/- 4.6% for the Hancock group and 88.1 +/- 4.2% for the Liotta group, and there was no significant difference in survival rate between the both groups. The actuarial survival rate at 14 years was 74.3 +/- 5.9% for the Hancock group. The valve-related complications in the Hancock and the Liotta groups were as follow; thromboembolism 1.9%/p-y vs 1.3%/p-y, bleeding 0 vs 0.7%/p-y, perivalvular leak 0.3% vs 0, infection 0.3%/p-y vs 0.7%/p-y, primary tissue failure (PTF) 3.3%/p-y and all valve-related complications 5.9%/p-y vs 6.0%/p-y. There was no significant difference in valve-related complications between the both groups. However, the actuarial event free rate of PTF was significantly lower in the Liotta group than the Hancock group between 4 and 7 years after operation (100% vs 88.8 +/- 4.3% in the 4th year p less than 0.01, 87.7 +/- 4.4% vs 73.5 +/- 8.0% in the 7th year, p less than 0.05). Although the both anti-thrombogenicity and anti-inflammation were acceptable in the porcine bioprostheses, this prosthesis is now used only in the selected patients because of the limited long term durability of this valve.