To evaluate the influence of antiplatelet drugs on patency in femoropopliteal vein bypasses, 48 vascular surgeons recruited 549 patients to a randomized double-blind trial of aspirin (300 mg) + dipyridamole (150 mg) or placebo twice daily starting 2 days before surgery and continuing indefinitely. Graft occlusion measured objectively by independent coordinators and cardiovascular events (myocardial infarction or stroke) were studied, expressed by life table, and analyzed statistically by log rank and confidence intervals (95% CI). Randomization achieved comparable groups with 60% of grafts inserted for rest pain or gangrene. Operative complications on aspirin plus dipyridamole included 18 reoperations for bleeding and 12 hematomas compared with 9 and 14, respectively, on placebo (NS). Most of the 172 graft failures occurred early with failure rates of 43/1000 patient-months in the first 3 months, reducing to 17/1000 at 6 to 12 months, and under 10/1000 in subsequent years. Cumulative graft patency on placebo was 72%, 62%, and 60% at 1, 2, and 3 years, respectively, compared with 78%, 70%, and 61% on aspirin plus dipyridamole. The difference in patency of 6.1% (95% CI, -3% to 15.5%) at 1 year and 8.0% (95% CI, -5% to 21%) at 2 years failed to achieve significance (p = 0.43). On mean follow-up of 34 months, 53 (132/1000 patient-years) cardiovascular events (myocardial infarction or cerebrovascular accident) occurred in patients on placebo compared with only 35 (73/1000) on aspirin plus dipyridamole, a significant difference of 59/1000 (p = 0.004). Antiplatelet therapy had little influence on femoropopliteal vein patency, but subsequent myocardial infarction and stroke was reduced in these patients with peripheral vascular disease.