Whereas adenosine itself exerted independent stimulatory and inhibitory effects on the adenylate cyclase activity of a platelet particulate fraction at low and high concentrations respectively, 2-substituted and N6-monosubstituted adenosines had stimulatory but greatly decreased inhibitory effects. Deoxyadenosines, on the other hand, had enhanced inhibitory but no stimulatory effects. The most potent inhibitors found were, in order of increasing activity, 9-(tetrahydro-2-furyl)adenine (SQ 22536), 2',5'-dideoxyadenosine and 2'-deoxyadenosine 3'-monophosphate. Kinetic studies on prostaglandin E1-activated adenylate cyclase showed that the inhibition caused by either 2',5'-dideoxyadenosine or compound SQ 22536 was non-competitive with MgATP and that the former compound, at least, showed negative co-operativity; 50% inhibition was observed with 4 micron-2',5'-dideoxyadenosine or 13 micron-SQ 22536. These two compounds also inhibited both the basal and prostaglandin E1-activated adenylate cyclase activities of intact platelets, when these were measured as the increases in cyclic [3H]AMP in platelets that had been labelled with [3H]adenine and were then incubated briefly with papaverine or papaverine and prostaglandin E1. Both compounds, but particularly 2',5'-dideoxyadenosine, markedly decreased the inhibition by prostaglandin E1 of platelet aggregation induced by ADP or [arginine]vasopressin as well as the associated increases in platelet cyclic AMP, so providing further evidence that the effects of prostaglandin E1 on platelet aggregation are mediated by cyclic AMP. 2'-Deoxyadenosine 3'-monophosphate did not affect the inhibition of aggregation by prostaglandin E1, suggesting that the site of action of deoxyadenosine derivatives on adenylate cyclase is intracellular. Neither 2',5'-dideoxyadenosine nor compound SQ 22536 alone induced platelet aggregation. Moreover, neither compound potentiated platelet aggregation or the platelet release reaction when suboptimal concentrations of ADP, [arginine]vasopressin, collagen or arachidonate were added to heparinized or citrated platelet-rich plasma in the absence of prostaglandin E1. These results show that cyclic AMP plays no significant role in the responses of platelets to aggregating agents in the absence of compounds that increase the platelet cyclic AMP concentration above the resting value.