P rostacyclin (PGI2) is an inhibitor of platelet function in vitro. We tested the hypothesis that PGI2 is formed in biologically active concentrations at the platelet-vascular interface in man and can be pharmacologically modulated to enhance its inhibitory properties. This became feasible when we developed a microquantitative technique that permits the measurement of eicosanoids in successive 40-microliters aliquots of whole blood emerging from a bleeding time wound. In 13 healthy volunteers the rate of production of thromboxane B2 (TXB2) gradually increased, reaching a maximum of 421 +/- 90 (mean +/- SEM) fg/microliters per s at 300 +/- 20 s. The hydration product of PGI2, 6-keto-PGF1 alpha, rose earlier and to a lesser degree, reaching a peak (68 +/- 34 fg/microliters per s) at 168 +/- 23 s. The generation of prostaglandins PGE2 and D2 resembled that of PGI2. Whereas the threshold concentration of PGI2 for an effect on platelets in vitro is approximately 30 fg/microliters, only less than 3 fg/microliters circulates under physiological conditions. By contrast, peak concentrations of 6-keto-PGF1 alpha obtained locally after vascular damage averaged 305 fg/microliters. Pharmacological regulation of PG endoperoxide metabolism at the platelet-vascular interface was demonstrated by administration of a TX synthase inhibitor. The rate of production of PGI2, PGE2, and PGD2 increased coincident with inhibition of TXA, as reflected by three indices; the concentration of TXB2 in bleeding time blood and serum, and excretion of the urinary metabolite, 2,3-dinor-TXB2. These studies indicate that PGI2 is formed locally in biologically effective concentrations at the site of vessel injury and provide direct evidence in support of transcellular metabolism of PG endoperoxides in man.