Adjuncts to thrombolytic agents have improved coronary patency and prevented early reocclusion after thrombolysis in acute myocardial infarction. The aim of this study was to compare in a canine thrombolysis model the effects of Ro 43-5054 = N-(N-[N-(p-amidinobenzoyl)-beta-alanyl]-L-alpha-aspartyl)-3- phenyl-L-alanine-trifluor-acetate, a new glycoprotein IIb-IIIa receptor antagonist with aspirin or heparin. Six groups of 10 dogs each were studied. A platelet-rich coronary thrombus was induced in open-chest dogs by electrical stimulation. In addition to recombinant tissue-type plasminogen activator (30 micrograms/kg.min during 60 min), the dogs received 1) saline, 2) heparin 200 U/kg + 50 U/kg.hr i.v., 3) aspirin 10 mg/kg i.v., 4) heparin+aspirin, 5) Ro 43-5054 (3 micrograms/kg.min) and 6) heparin+Ro 43-5054. The overall reperfusion rate was 70% (range, 60-90%) and comparable in all the six groups. During the 120-min observation period, episodes of reocclusion were observed in the absence of antiplatelet therapy (group 1 and 2) irrespective of heparin treatment. Aspirin prevented coronary reocclusion in half of the reperfused dogs (group 3 and 4). However, after reinforcement of the thrombogenic stimulus, 80% of the reperfused dogs treated with aspirin showed reocclusion, whereas none of them reoccluded when treated with Ro 43-5054. Thus, inhibition of platelet activation by the selective, nonpeptidic glycoprotein IIb-IIIa receptor antagonist Ro 43-5054, although without effect on the time to reperfusion, better protected than aspirin against early reocclusion after thrombolytic therapy.