Thromboxane A2, the predominant cyclooxygenase product in platelets, is a potent platelet agonist and vasoconstrictor in vitro. Prostacyclin, the major product of vascular endothelium, has opposite effects on platelet function and vascular tone. These properties prompted the hypothesis that a "balance" between these compounds regulated interactions between platelets and the vessel wall in vivo. Although this possibility has been addressed extensively through experiments in vitro, clinical investigations commonly have been confounded by problems with analytic methodology, by selection of inappropriate metabolic targets for analysis, and by artifacts of trial design. The most reliable forms of assessing biosynthesis that are currently available still do not provide definitive information as to the tissue of origin of the compound studied and are directed toward stable but biologically inactive metabolites rather than the evanescent primary compounds themselves. Despite these limitations, both biochemical evidence and clinical trials clearly implicate thromboxane A2 as an important mediator of vascular occlusive disease in humans. The role of prostacyclin is much more conjectural. It does not circulate in concentrations sufficient to exert a systemic effect, but it may play a local homeostatic role in the regulation of platelet-vascular interactions. Whether preservation of the capacity to form prostacyclin coincident with inhibition of thromboxane A2 is of functional importance can be addressed only by clinical trials comparing inhibitors of thromboxane synthesis inhibition that are selective with cyclooxygenase inhibitors that also block the biosynthesis of prostacyclin. The recognition that multiple factors have the potential to regulate both platelet and vascular function at their interface renders the concept of a thromboxane A2-prostacyclin "balance" somewhat unlikely. However, both eicosanoids may interact with other factors to determine the development or persistence of vascular occlusion. Inhibition of the synthesis or function of thromboxane A2 remains the predominant mechanism for achieving interference with platelet function in vivo. Accumulating evidence for the efficacy of aspirin in human syndromes of vascular occlusion suggests that the biologic role of these compounds in humans should be pursued.