To assess the optimal level of oral anticoagulation to prevent occlusion of vein coronary bypass grafts, 318 patients from a graft patency trial were analysed retrospectively. Oral anticoagulant therapy was started one day before surgery and continued for one year, after which graft occlusion was assessed by angiography. The aimed level of anticoagulation was 2.8-4.8 International Normalized Ratio (INR). Clinical outcome was assessed by the incidence of myocardial infarction, thrombosis and major bleeding. The observed anticoagulation level was 2.8-4.8 INR for 54%, and 1.8-3.8 INR for 75% of time per patient. Occlusion rates in patients who spent < 35, 35-70, and > or = 70% of time within INR range 2.8-4.8 were 10.5%, 10.8% and 11.8%, respectively (differences not statistically significant). Patients who spent > or = 70% of time within INR range 1.8-3.8 versus 2.8-4.8 showed comparable occlusion rates. The risk of graft occlusion was not related to quality of anticoagulation early (0-3 months) or late (3-12 months) after surgery. Myocardial infarction, thrombosis and major bleeding occurred in 1.3%, 2.0% and 2.9% of patients. To maintain vein graft patency in the first postoperative year by oral anticoagulation, a level within INR range 1.8-3.8 for > or = 70% of time seems to be sufficient.