The St. Jude Medical valve is a bileaflet prosthesis with excellent hemodynamic characteristics, but the long-term surgical experience with this valve, its durability, and its biocompatibility are unknown. During a 10-year period from March 1978 to 1988, 690 prostheses (290 aortic, 252 mitral, and 74 double aortic-mitral) were inserted as the initial valve replacement substitute in 616 patients (mean age 63 years). Coronary atherosclerosis was present in 58%. Follow-up totaled 2031 patient-years (mean 3.3 years) and was 95% complete (32 lost). Early (30-day) mortality rates were 5.2%, 11.9%, and 8.1% after aortic, mitral, and double valve replacement; 5- and 9-year actuarial survival rates were 71% +/- 3% and 51% +/- 8%, 59% +/- 4% and 41% +/- 6%, and 69% +/- 6% and 47% +/- 15%, respectively. Deaths were associated with extensive coronary atherosclerosis (p less than 0.001), older age (p less than 0.001), advanced preoperative New York Heart Association functional class (p less than 0.05), and malignant ventricular arrhythmias (p less than 0.05). No structural failures have been observed. Embolism (40 events) occurred at a rate of 2.0%/pt-yr (2.3% aortic, 1.6% mitral, 2.0% double). There were six cases of valve thrombosis (0.3%/pt-yr; one fatal). Hemorrhage was the most frequent complication (2.6%/pt-yr); 13 (25%) of 52 events were fatal, accounting for 62% of all valve-related deaths. After the target prothrombin time ratio was lowered, the rate of hemorrhage decreased by 44% (2.7% to 1.5%/pt-yr), while the combined rate of embolism and valve thrombosis increased slightly (2.2% to 2.5%/pt-yr, a 14% change). In summary, the St. Jude Medical valve remains a durable valve substitute. Survival was strongly related to the presence of associated coronary atherosclerosis. The most common complication has been hemorrhage; a less intensive warfarin regimen may reduce hemorrhagic risk while maintaining thromboembolic protection.