Our laboratory has demonstrated previously that prostaglandins are partially responsible for the vasodilation of rat cremaster muscle arterioles in vivo to a brief occlusion or hydrogen peroxide (H2O2). In the present study, in pentobarbital-anesthetized rats, we investigated the mechanism of the prostaglandin-independent portion of the dilation to these stimuli by measurement of changes in the diameter of third order cremasteric arterioles (approximately 15 microns) by video microscopy. In the presence of indomethacin suffusion (10 micrograms/ml), arteriolar dilation to the release of a 15-sec occlusion of a single arteriole or to topical application of 0.1 to 1 mM H2O2 (100 microliters) was reduced significantly by suffusion of 20 microM quinacrine or 5 microM methylene blue, whereas vasodilation to adenosine was not affected by these probes. Furthermore, the reactive hyperemia was not altered by suffusion of 50 microM hydroquinone or 0.2 mM NG-monomethyl-L-arginine, inhibitors of the dilation to acetylcholine mediated by the endothelium-derived relaxing factor. Reactive hyperemia was also not affected by 30 microM theophylline (an adenosine antagonist) or suffusion of catalase (160 U/ml) plus superoxide dismutase (240 U/ml). Therefore, the reactive hyperemia does not appear to be mediated through endothelium-derived relaxing factor, adenosine or the extracellular formation of reactive O2 species. However, these observations suggest that the prostaglandin-independent vasodilation to H2O2 and to the release of a brief arteriolar occlusion may be mediated by a common mechanism, possibly involving cyclic GMP. Similarities in pharmacological modulation of the dilation after occlusion and to H2O2 in the cremaster microcirculation suggest consideration of the involvement of H2O2 in the hyperemic response.