The present data disagree with earlier suggestions that thrombin's effect on platelets is to cause a decrease in intracellular cyclic 3',5'-adenosine monophosphate. Washed human platelets or platelet-rich plasma were incubated at 37 degrees C with human thrombin. After centrifugation, the supernates were assayed for nucleotides and calcium released. The platelet pellets, and in some experiments the supernates as well, were assayed by radioimmunoassay for intracellular cyclic AMP. In the washed platelet system, increasing doses of thrombin to 0.5 U/cc induced increasing release of nucleotides and calcium. This was accompanied by an average twofold increase in intracellular cyclic AMP levels. Prostaglandin E(1), which inhibited 30-50% of release, induced a four- to fivefold increase in cyclic AMP levels that was additive to the cyclic AMP-stimulatory effect of thrombin. Theophylline, which inhibited only 20-40% of nucleotide release, was synergistic with thrombin in the intracellular accumulation of cyclic AMP. The time-course of cyclic AMP accumulation in response to thrombin was slower than thrombin-induced nucleotide release. Similar findings were made in the platelet-rich plasma system where thrombin stimulation of nucleotide release also resulted in a marked accumulation of intracellular cyclic AMP. Thrombin did not appear to stimulate the release of intracellular cyclic AMP. The mechanism underlying these observations was not apparent. The thrombin had no measurable inhibitory effect on platelet phosphodiesterase activity in either intact washed cells or the platelet homogenate supernates. Furthermore, thrombin inhibited, rather than stimulated, platelet adenyl cyclase activity in both intact washed cells and washed platelet particulate fractions. Of note, however, was the finding that thrombin did not completely inhibit the adenyl cyclase activity of prostaglandin-stimulated cells. Further work is needed to clarify the significance of this observation.Nonetheless, the accumulation of intracellular cyclic AMP in response to thrombin observed in the present study suggests that the antagonistic actions of various agents on the platelet release reaction, thought to underlie platelet function, may depend upon a mechanism more intricate than a straightforward mediation through directly opposite effects on platelet cyclic AMP.