Experimental animal models for erythrocyte-rich (ER) and platelet-rich (PR) arterial thrombosis were developed in dogs and rabbits and used for the evaluation of the effect of antithrombin and antiplatelet agents on thrombolysis with recombinant tissue-type plasminogen activators (rt-PA). The canine models consist of a whole blood clot produced in the left anterior descending coronary artery (ER thrombus) or a 1-cm everted (inside-out) segment graft in the circumflex coronary artery that predisposes to occlusion with PR material (PR thrombus). The rabbit models consist of a femoral arterial whole blood clot (ER thrombus) or a femoral arterial eversion graft (PR thrombus). The whole blood clot models are sensitive to recanalization with rt-PA but are consistently associated with reocclusion, notwithstanding the concomitant use of heparin and/or aspirin. Clot lysis is accelerated and reocclusion is prevented by the administration of F(ab')2 fragments of a monoclonal antibody 7E3 directed against the platelet glycoprotein IIb/IIIa receptor; of Argatroban, a synthetic thrombin inhibitor; or of kistrin, a glycoprotein IIb/IIIa-blocking polypeptide from the Malayan pit viper venom. The PR thrombus models are very resistant to recanalization with rt-PA, but this resistance can be overcome by the concomitant use of the platelet glycoprotein IIb/IIIa-blocking antibody. Thus, selective platelet glycoprotein IIb/IIIa inhibitors are more effective than aspirin, heparin, or both in accelerating arterial thrombolysis with rt-PA; in preventing reocclusion after clot lysis; and in overcoming the resistance of PR thrombus to dispersion with rt-PA. These experimental animal models may be useful in the development of improved thrombolytic strategies using plasminogen activators in conjunction with specifically targeted antiplatelet and anticoagulant agents.