Retrospectively, we reviewed the charts of 101 patients at the University of Kansas Medical Center who received low-intensity anticoagulation for mechanical prosthetic valves implanted over a 17-yr period. The mean duration of follow-up was 4.6 yr, and the total duration of follow-up was 466.5 patient-yr. The patients' records were evaluated for evidence of hemorrhagic or thromboembolic complications. A prothrombin time ratio of 1.3 to 1.5 times control was considered to be low-intensity anticoagulation. There were three thromboembolic events or 2.9/100 patient-yr of follow-up at a prothrombin time ratio of less than 1.3, four thromboembolic events or 2.5/100 patient-yr of follow-up at 1.3 to 1.5 times control, four thromboembolic events or 2.2/100 patient-yr of follow-up at 1.6 to 2.0 times control, and no thromboembolic events at prothrombin time ratios greater than 2.0 times control. Hemorrhagic events occurred in three patients at a prothrombin time ratio of less than 1.3 times control or 2.8/100 patient-yr of follow-up, in six patients at 1.3 to 1.5 times control or 3.8/100 patient-yr of follow-up, in ten patients at 1.6 to 2.0 times control or 5.5/100 patient-yr of followup, and in two patients at 2.1 to 2.5 times control or 12.2/100 patient-yr of follow-up. The rate of hemorrhagic events at 2.5 times control was 470/100 patient-yr follow-up. While not providing definitive proof, we believe that our retrospective study provides supportive evidence for the use of low-intensity anticoagulation in patients with mechanical cardiac prostheses.