Addition of the 3-series fatty acid precursor (icosapentaenoic acid, IPA), its endoperoxide [prostaglandin (PG)H(3)], or thromboxane A(3) to human platelet-rich plasma (PRP) does not result in aggregation of the platelets. In fact, preincubation of human PRP with exogenous PGH(3) actually inhibited aggregation by increasing platelet cyclic AMP concentrations. PGH(3) undergoes rapid spontaneous degradation to PGD(3) in human PRP. The PGD(3) so formed is adequate to account for the increase of platelet cAMP and inhibition of aggregation. Furthermore, addition of PGD-specific antisera to human PRP blocked the platelet inhibitory activity of exogenous PGH(3). PGD(3) has considerable potential as a circulating antithrombotic agent. Pretreatment of human PRP with the adenylate cyclase inhibitor 2',5'-dideoxyadenosine blocked the increase of platelet cyclic AMP and the inhibition of aggregation normally produced by PGI(2), PGE(1), PGD(2), PGH(3), and PGD(3). Furthermore, the dideoxyadenosine unmasked a direct but moderate reversible aggregatory effect in response to the subsequent addition of PGH(3). Similarly, the dideoxyadenosine markedly enhanced the aggregation produced by exogenous PGH(2). IPA is readily incorporated into tissue lipids but proved to be a poor substrate for kidney, blood vessel, or heart cyclooxygenase. IPA was previously shown to be a poor substrate for platelet cyclooxygenase. IPA is readily deacylated from the renal phospholipid pool in response to bradykinin, a substance that also stimulates the release of arachidonic acid. A diet that relies primarily on cold-water fish, as in the case of the Greenland Eskimos, lowers endogenous arachidonic acid and markedly increases the IPA content of tissue lipids. Thus, because IPA has the potential to act as an antagonist with arachidonic acid for platelet cyclooxygenase and lipoxygenase, the simultaneous release of IPA could suppress any residual arachidonic acid conversion to its aggregatory metabolites.