From August 1976 to January 1980, 135 patients, age 20 years or younger (mean 13.7 +/- 3.46 years), received 37 Hancock, 4 Angell-Shiley and 94 Carpentier-Edwards prostheses as a mitral valve substitute. Only patients who had isolated mitral valve replacement were included in this study. All patients were in NYHA class III and IV and 22 had acute rheumatic valvulitis. Approximately half of the patients had mitral regurgitation. Early mortality was 4.4% (6 patients). The cumulative follow up period was 356 patient-years. Thirty-eight patients died late (10.7% per patient year), of which 26 (7.3% per patient year) died of valve related causes. Sixty-four patients (18% per patient year) had mitral valve replacement for a degenerated bioprosthesis. Eleven patients with degenerating prostheses died without surgery and 5 await reoperation. Hence, structural valve failure occurred in 80 instances (22.5% per patient year). Actuarial survival at 7 years was 50%. Complication-free valve survival was 15%. Only 7 patients (5.2%) remain with their original prostheses, free from degeneration. Our experience confirms that bioprostheses have a prohibitive failure rate in children. Degenerating valves should be replaced early rather than late.