We have investigated the effects of L-NG-nitro arginine (L-NOARG), glibenclamide, ouabain, tetraethylammonium and 4-aminopyridine on the methacholine-induced endothelium-dependent vasodilation in perfused resistance arteries from spontaneously hypertensive rats (SHR) and Wistar Kyoto rats (WKY). Since the concentration-response curves of MCh were similar in both types of preparations there does not seem to exist an endothelial dysfunction in mesenteric arteries of SHR. L-NOARG only partially inhibited the maximal methacholine-induced response in preparations taken from SHR and WKY rats. Ouabain decreased the maximal effect of methacholine without altering the potency (pD2). Preparations from SHR were more susceptible to ouabain. 4-aminopyridine and tetraethylammonium decreased the pD2 for methacholine without reducing the maximal effect (Emax). The WKY rat preparations were more affected by these compounds. An important role of ATP-sensitive potassium channels may be ruled out since glibenclamide was without effect on the methacholine-induced vasodilation. It is concluded that endothelium-derived relaxing factor is only partially responsible for the endothelium-dependent vasodilation. Indirect arguments point toward a role of endothelium-derived hyperpolarizing factor, since ouabain, tetraethylammonium and 4-aminopyridine inhibited the methacholine-induced response. Although hypertension related differences for these compounds were observed high blood pressure does not seem to alter the functional response to muscarinic stimulation.