W e have evaluated the contributions of nitric oxide (NO) and prostacyclin (PGI2) in the in vivo antiplatelet effects of clinically useful nitrovasodilators. In rats, intravenous infusion of three NO donors, glyceryl trinitrate, sodium nitroprusside, or 3'-morpholinosydnonimine, the stable metabolite of molsidomine, released 6-keto PGF1alpha (the stable metabolite of PGI2) and inhibited ex vivo human platelet aggregation to adenosine diphosphate by at least 80%. In in vitro studies, glyceryl trinitrate, sodium nitroprusside, and 3'-morpholinosydnonimine, at clinically attainable concentrations, increased cyclooxygenase activity in endothelial cells (EC), which resulted in a four- to sixfold release of 6-keto PGF1alpha. Pretreatment of the EC with hemoglobin which binds to and inactivates the biological actions of NO, but not by methylene blue (MelB), attenuated the NO-mediated PGI2 from the EC by at least 70%. Release of 6-keto PGF1alpha by the NO donors increased the ability of these compounds to inhibit thrombin-induced human platelet aggregation by at least 10 times; this potentiation was inhibited by hemoglobin but not by MeB. MeB blocked the direct anti-platelet effect of the NO donors in the absence of EC. In summary, we have demonstrated that NO, directly as well as together with an NO-driven cyclooxygenase activation (and hence PGI2), release contributes to the marked anti-platelet effects observed after the in vivo administration of clinically used nitrovasodilators.