EDRF (endothelium-derived relaxing factor) is a cellular and intercellular messenger that activates soluble guanylate cyclase. In blood vessels it is released from the endothelium and causes relaxation of vascular smooth muscle. Halothane previously has been shown to attenuate EDRF-induced vasodilation elicited by the receptor-mediated vasodilators acetylcholine and bradykinin and to alter muscarinic receptor activity. We examined and compared the effects of the inhaled anesthetics halothane, enflurane, and isoflurane on endothelium-dependent vasodilation and tested the hypothesis that these agents inhibit EDRF-mediated vasodilation solely through inhibition of endothelial cell receptor-mediated EDRF release. Isolated rat thoracic aortic rings were mounted for isometric tension recording and preconstricted with phenylephrine. Cumulative dose-response curves were obtained to methacholine, a receptor-mediated endothelium-dependent dilator; to A23187, a nonreceptor-mediated endothelium-dependent dilator; and to sodium nitroprusside, a direct-acting endothelium-independent dilator before, during, and after inhalational anesthetic exposure. Both receptor-mediated and non-receptor-mediated endothelium-dependent relaxation by methacholine and A23187, respectively, were significantly (P less than 0.01 to P less than 0.05) and reversibly attenuated by halothane, enflurane, and isoflurane at 2 MAC and by isoflurane at 1 MAC. Endothelium-independent relaxation by sodium nitroprusside, an agent that acts directly on the vascular smooth muscle cell to activate guanylate cyclase, was unaffected by any of the anesthetics at any concentration tested. Indomethacin had no significant effect on the inhibition of endothelium-dependent vasodilation by these inhalational anesthetics. We conclude that halothane, enflurane, and isoflurane inhibit endothelium-dependent vasodilation; that isoflurane is more potent than halothane and enflurane in this regard.(ABSTRACT TRUNCATED AT 250 WORDS)