1. Non-nitric oxide/prostaglandin-mediated endothelium-derived hyperpolarization (EDH) is considered to be mediated, in part, by gap junctions and it has been suggested that cAMP increases endothelium-derived hyperpolarizing factor (EDHF)-mediated relaxation through the modulation of gap junctions. Cilostamide, which inhibits phosphodiesterase III, has been suggested to augment EDHF-type relaxation by increasing the concentration of cAMP. 2. In the present study, we investigated the effect of cilostamide on EDH per se in mesenteric arteries of Wistar rats using a conventional microelectrode technique. 3. The resting membrane potential of the mesenteric arteries was significantly more negative in the presence of 10(-6) mol/L cilostamide compared with control conditions. Furthermore, EDH in response to 10(-6) mol/L acetylcholine (ACh) in the presence of 10(-5) mol/L indomethacin and 10(-4) mol/L N(G)-nitro-L-arginine was decreased in the presence of 10(-6) mol/L cilostamide by approximately 5 and 3.5 mV in proximal and distal arteries, respectively. 4. Glibenclamide (10(-5) mol/L), an ATP-sensitive potassium channel (K(ATP)) inhibitor, abolished the hyperpolarization to 10(-6) mol/L cilostamide. Furthermore, in the presence of glibenclamide, ACh-induced EDH was unaffected by cilostamide, suggesting that the inhibition of ACh-induced hyperpolarization by cilostamide in the absence of glibenclamide may be due to the smaller driving force for hyperpolarization because of the more negative membrane potential under such conditions. 5. The findings of the present study suggest that cilostamide produces hyperpolarization by activating K(ATP) channels, presumably by increasing cAMP. However, cilostamide alone may not directly affect EDH.