OBJECTIVE This study was designed to evaluate the efficacy of local versus systemic treatment of thrombosis with various antithrombotic drugs. BACKGROUND Local use of low dose antithrombotic drugs has been proposed as being effective and safe. METHODS Heparin (30 U/kg), an antithrombin agent (argatroban, 0.05 mg/kg body weight) or a defibrinogenating drug (batroxobin, 0.05 U/kg) was locally infused into one side of the canine iliac artery after injury by balloon inflation. The other side was injured as a control. The efficacy of systemic delivery of high dose (heparin [300 U/kg] and argatroban [0.5 mg/kg]) and low dose drugs was also assessed. RESULTS Sixty minutes after local treatment in 22 dogs, no thrombotic stenosis was observed by angiography in locally treated arteries (p < 0.005 vs. mean thrombotic stenosis of 27% in control segments for heparin, 25.3% in control segments for argatroban and 32% in control segments for batroxobin). Angioscopy demonstrated the same trend. In locally treated arteries, thrombus weight was significantly lower in the treated than control side. In the systemic high dose group (n = 10), angiographic thrombotic stenosis was < 5% after high dose drug delivery (p < 0.05 vs. control segments, 37.4% for heparin, 43% for argatroban). In another 10 dogs, low dose systemic delivery was not effective in inhibiting thrombus formation. Activated partial thromboplastin time and fibrinogen levels did not change with local treatment. CONCLUSIONS Compared with systemic administration of antithrombotic drugs, local treatment is a safer and more effective method of preventing thrombosis.