Arterial thrombosis in the rabbit was established as a novel model to evaluate the effects of antithrombotic agents. Endothelial injury was produced by applying electrical stimulation to the femoral artery. The process of primary endothelial injury, and subsequent platelet activation and fibrin formation were confirmed by electron microscopy. In this model, vessel occlusion occurred within 30 min after stimulation without changes in heart rate and blood pressure. Using this model, several agents were evaluated for their antithrombotic activities: aspirin (30 mg/kg, p.o.), ticlopidine (10–100 mg/kg, p.o.), heparin (300 unit/kg, i.v.), PPACK (10–33 μg/kg/min, i.v.), WEB-2347 (1 mg/kg, p.o.) and nicardipine (10 μg/kg, i.v.). Fifty per cent decrease in vessel temperature (Tp, assessed by a thermic probe, averaged 15.1 ± 1.2 (n= 11, p.o.) and 15.6 ± 1.9 min (n= 8, i.v.) in the vehicle groups, and this was significantly prolonged by aspirin (23.0 ± 2.6 min), ticlopidine at a dose of 100 mg/kg (24.6 ± 2.5 min), heparin (27.2 ± 2.8 min) and PPACK at a dose of 33 μg/kg (30.0 min). However, WEB-2347 and nicardipine were without effect. The effect of aprosulate, a new class of polyanion with anticoagulant activity, was further examined. Aprosulate (1–30 mg/kg, i.v.) inhibited thrombus formation in a dose-dependent manner. These results show that acute occlusive thrombus can be readily and reproducibly formed in the rabbit femoral artery and suggest that this thrombus formation depends on the activation of both platelets and blood coagulation. The merit of this model lies in its simplicity for evaluating the antithrombotic effects of antiplatelet and anticoagulant agents and is therefore expected to be extensively used in the future.