The effect of heparin and of the synthetic competitive thrombin inhibitor (2R,4R)-4-methyl-1-[N2-(3-methyl-1,2,3,4-tetrahydro-8-quinolinesulfon yl)-L-arginyl]-2-piperidinecarboxylic acid monohydrate (argatroban) on platelet-rich arterial thrombosis was studied in a rabbit model, consisting of a 4-6-mm everted ("inside-out") femoral arterial segment. Intravenous injection of heparin (200 units/kg) failed to prevent occlusion within 60 minutes in all 10 rabbits, whereas intravenous argatroban infusion at a rate of 100 or 200 micrograms/kg/min for 60 minutes, which prolonged the thrombin time more than fourfold, prevented thrombosis in nine of 13 rabbits (p = 0.002 vs. i.v. heparin). Intra-arterial infusion of 200 units/kg heparin over 60 minutes prevented occlusion in six of nine rabbits (p = 0.003 vs. i.v. heparin), whereas intra-arterial argatroban at a rate of 100 micrograms/kg/min for 60 minutes prevented thrombosis in all 10 rabbits (p = 0.00001 vs. i.v. heparin). Patency of femoral arterial segments was maintained after the end of the intra-arterial heparin and intravenous or intra-arterial argatroban infusion for up to 3 hours despite normalization of the thrombin time and partial thromboplastin time. Pathologic examination of the graft revealed that the inverted adventitial surface was covered by layers of platelets without platelet aggregation or fibrin deposition. These findings indicate that thrombin plays an important role in platelet-rich arterial thrombosis, and that the thrombogenic stimulus is rapidly attenuated by short-term infusion of the synthetic thrombin inhibitor. Selective thrombin inhibition can constitute an alternative approach to the prevention of arterial occlusion after angioplasty or thrombolytic therapy in patients with unstable coronary syndromes.