It is unclear to what extent the endothelium-derived hyperpolarizing factor (EDHF) contributes to the control of microcirculatory blood flow in vivo. We analyzed, by intravital microscopy in hamster muscles, the potential role of EDRF along the vascular tree under stimulated (ACh) or basal conditions. Experiments were performed in conscious as well as anesthetized (pentobarbital, urethan) animals. Additionally, cellular effects of the potential EDHF were studied in isolated small arteries. In pentobarbital-anesthetized animals, treatment with N-omega-nitro-L-arginine. (L-NNA; 30 mu mol/l) and indomethacin (3 mu mol/l) reduced the dilation in response to 10 mu mol/l ACh from 60 +/- 6 to 20 +/- 4%. This nitric oxide/prostaglandin-independent dilation (NPID), which was of a similar magnitude in large and small arterioles, was abolished by potassium depolarization or charybdotoxin (ChTX, 1 mu mol/l) but not by glibenclamide. In conscious animals, NPID amounted to 33 +/- 3%. The inhibitor of the P-450 monooxygenase 17-octadecynoic acid (ODYA) reduced NPID further to 9 +/- 4%. ChTX abolished the NPID and also reduced basal diameters (by -11 +/- 3%). The induction of anesthesia with pentobarbital reduced NPID (to 12 +/- 6%), whereas urethan anesthesia was without effect. Pentobarbital also reduced the ACh-induced hyperpolarization of vascular smooth muscle ih isolated arteries, whereas ChTX abolished it. This study suggests that a considerable part of the ACI 1 dilation in the microcirculation is mediated by EDHF, which also contributes to the control of basal tone in conscious animals. The direct inhibitory effect of pentobarbital and ODYA supports the idea that "microcirculatory" EDHF is a product of the cytochrome P-450 pathway. The role of EDHF might be underestimated in pentobarbital-anesthetized animals.