We evaluated 643 patients surviving aortic, mitral, and aortic and mitral ("double") valve replacement with the Björk-Shiley prosthesis from 1 to 72 months (median 38 months) postoperatively. Intermediate-term survival rate was similar to that reported for other prosthetic and bioprosthetic devices. Factors unrelated to the device, but related to preoperative patient characteristics or intraoperative or early postoperative events, had important association with late survival rate. These included previous valve replacement, concomitant ventricular aneurysm resection, black race, preoperative functional class, method of myocardial protection, duration of ischemic arrest, ventricular arrhythmias, and double valve replacement. Concomitant coronary artery bypass grafting and the type of valve lesion had no effect on survival prognosis. Freedom from thromboembolism was equal to or superior to other devices, but valve thrombosis occurred 16 times in 15 patients and caused 13 deaths. Improvement in New York Heart Association (NYHA) functional class occurred in the majority of patients. We conclude that the Björk-Shiley valve is durable and effective, but, as with other devices, is associated with problems of thromboembolism and thrombosis. Intermediate-term survival is related also to non-device, patient-associated characteristics.