Objective: To assess the performance of the Medtronic Hall valve (Medtronic, Inc, Minneapolis, Minn) in one institution over a 20-year period. Methods: Since 1979, Medtronic Hall valves have been used in 1766 procedures (736 aortic, 796 mitral, and 234 double). Patients were followed up prospectively at 6- to 12-month intervals for a total of 12,688 follow-up years. Anticoagulation data (international normalized ratio) were recorded for all patients (approximately 95,000 observations). Results: Linearized rates of valve-related late death for aortic, mitral, and double valve replacement were 0.8%/y, 0.9%/y, and 1.1%/y, respectively. Risk factors for late mortality were (relative risk) diabetes (1.9), decade of age (1.6), concomitant coronary artery bypass grafting (1.4), hypertension (1.3), non-sinus rhythm (1.3), large valve size (1.1), valve regurgitation (1.3), and male sex (1.2). For aortic, mitral, and double valve replacement, linearized rates (percent per year) of adverse events were valve thrombosis 0.04, 0.03, and 0.0; all thromboembolism 2.3, 4.0, and 3.4; stroke 0.6, 0.8, and 0.6; major hemorrhage 1.2, 1.4, and 1.6; and prosthetic endocarditis 0.4, 0.4, and 0.7. Risk factors for thromboembolism were (relative risk) mitral valve replacement (1.9), diabetes (1.8), hypertension (1.5), and history of embolism (1.4). Conclusion: At 20 years the Medtronic Hall valve demonstrates excellent durability, good hemodynamic performance, and very low thrombogenicity, with a valve thrombosis rate lower than those reported for bileaflet designs. With this prosthesis, both survival and thromboembolic events are predominantly determined by patient risk factors.(J Thorac Cardiovasc Surg 2001;121:1090-100)