Reactive oxygen metabolites have been reported to affect platelet aggregation. However, this phenomenon is still poorly understood. In the present study we investigated the effects of superoxide radical and hydrogen peroxide (H2O2) on platelet function in vitro and correlated those effects to possible changes of platelet concentrations of cyclic nucleotides and thromboxane, since these systems play a key role in the response of platelets to activating stimuli. Human platelets were exposed to xanthine-xanthine oxidase (X-XO), a system that generates both superoxide radicals and H2O2. Sixty seconds of incubation with X-XO impaired aggregation in response to ADP (by 48%), collagen (by 71%), or the thromboxane mimetic U-46619 (by 50%). This effect was reversible and occurred in the absence of cell damage. Impairment of aggregation in platelets exposed to X-XO was due to H2O2 formation, since it was prevented by catalase but not by superoxide dismutase. Similarly, incubation with the pure H2O2 generator glucose-glucose oxidase also markedly inhibited ADP-induced platelet aggregation in a dose-dependent fashion. Impaired aggregation by H2O2 was accompanied by a > 10-fold increase in platelet concentrations of guanosine 3',5'-cyclic monophosphate (cGMP), whereas adenosine 3',5'-cyclic monophosphate levels remained unchanged. The inhibitory role of increased cGMP formation was confirmed by the finding that H2O2-induced impairment of platelet aggregation was largely abolished when guanylate cyclase activation was prevented by incubating platelets with the guanylate cyclase inhibitor, LY-83583. Different effects were observed when arachidonic acid was used to stimulate platelets. Exposure to a source of H2O2 did not affect aggregation to arachidonate. Furthermore, in the absence of exogenous H2O2, incubation with catalase, which had no effects on platelet response to ADP, collagen, or U-46619, virtually abolished platelet aggregation and markedly reduced thromboxane B2 production (to 44% of control) when arachidonic acid was used as a stimulus. In conclusion, our data demonstrate that H2O2 may exert complex effects on platelet function in vitro. Low levels of endogenous H2O2 seem to be required to promote thromboxane synthesis and aggregation in response to arachidonic acid. In contrast, exposure to larger (but not toxic) concentrations of exogenous H2O2 may inhibit aggregation to several agonists via stimulation of guanylate cyclase and increased cGMP formation.