To resolve questions of drug actions, efficacy, and interactions for platelet-modifying agents used clinically, we have compared the relative capacities and mechanisms of aspirin, dipyridamole, sulfinpyrazone, and dazoxiben to prevent arterial thromboembolism in a baboon model. In 136 studies the agents were given twice daily by oral administration both singly and in combination. The antithrombotic efficacy of a given therapy was determined by its capacity to interrupt steady-state platelet utilization induced by thrombogenic arteriovenous cannulae. When given alone, dipyridamole and sulfinpyrazone reduced the rate at which platelets were utilized by thrombus formation in a dose-dependent manner with essentially complete interruption by dipyridamole at 10 mg/kg per d. In contrast, neither aspirin (2-100 mg/kg per d) nor dazoxiben (20-100 mg/kg per d) decreased cannula platelet consumption detectably despite the striking reduction in the capacity of platelets to produce thromboxane B2. However, aspirin, but not dazoxiben, potentiated the antithrombotic effects of dipyridamole and sulfinpyrazone in a dose-dependent fashion without changing the pharmacokinetics for any of the agents. Complete potentiation required aspirin at 20 mg/kg per d to be given with each dose of dipyridamole. Because dazoxiben's blockade of platelet thromboxane A2 production was not associated with antithrombotic potentiation, and because complete potentiation by aspirin required a dose that fully inhibited vascular production of prostaglandin I2 (PGI2), we conclude that aspirin's potentiating effect on dipyridamole is independent of PGI2 production or inhibition of thromboxane A2 formation. In addition, because frequent repeated and synchronous dosing of aspirin was necessary, aspirin's potentiating effects appear to be produced by mechanism(s) unrelated to its potent, irreversible inhibition of platelet cyclooxygenase.