From 1977 to 1987, 829 Ionescu-Shiley pericardial valves (Shiley, Inc., Irvine, Calif.) were implanted in 766 patients at the University of Ottawa Heart Institute. There were 476 patients who had aortic valve replacement, 234 who had mitral valve replacement, and 44 who had double valve replacement. The standard-profile design was used in 508 patients and the low-profile design in 321 patients. Follow-up was obtained for 97% of patients, with calculation of event-free probabilities. At 10 years the overall probability of freedom from structural failure was 48% +/- 7% after aortic valve replacement, 44% +/- 15% after mitral valve replacement, and 79% +/- 11% after double valve replacement. Although at 5 years the probability of failure was statistically lower with the low-profile design, this favorability was lost by 6 years. Freedom from structural failure was only 47% +/- 7% for the standard-profile valve at 10 years. Thus the probability of freedom from reoperation was only 46% +/- 7% after aortic valve replacement, 39% +/- 6% after mitral valve replacement, and 65% +/- 20% after double valve replacement at 10 years. Thromboembolism occurred in 69 patients, for a predicted freedom from this complication at 10 years of 79% +/- 3% after aortic, 73% +/- 7% after mitral, and 96% +/- 4% after double valve replacement. There were 31 cases of endocarditis. The 10-year predicted freedom from endocarditis, therefore, was 86% +/- 3% after aortic, 98% +/- 1% after mitral, and 97% +/- 1% after double valve replacement. A total of 221 operative and late deaths were recorded in this series. Prosthetic valve failure accounted for 27% of late deaths. The 10-year survival rates were estimated to be 56% +/- 5% (aortic valve replacement), 54% +/- 6% (mitral valve replacement), and 51% +/- 8% (double valve replacement). We concluded that the Ionescu-Shiley pericardial xenograft provides less than optimal clinical performance and its use has been discontinued.