In the glycerol model of renal injury we describe an acute rise in systemic arterial pressure which is attended by a reduced vasodilatory response to acetylcholine in vivo; vasodilatory responses to verapamil, however, were not impaired. Neither arginine nor sodium nitroprusside diminished this rise in blood pressure; N-nitro-L-arginine methyl ester (L-NAME) elevated basal mean arterial pressure and markedly blunted the rise in mean arterial pressure following the administration of glycerol. Aortic rings from the glycerol-treated rat demonstrate an impaired vasodilatory response to acetylcholine, an effect not repaired by arginine; the vasodilatory responses to nitric oxide donors, sodium nitroprusside and SIN-1, were also impaired; 8-bromo-cGMP, at higher doses, evinced a vasodilatory response comparable to that observed in the control rings. This pattern of responses was not a nonspecific effect of aortic injury, since aortic rings treated with mercuric chloride, a potent oxidant, displayed an impaired vasodilatory response to acetylcholine but not to sodium nitroprusside. We conclude that in the glycerol model of heme protein-induced tissue injury, there is an acute elevation in mean arterial pressure attended by impaired endothelium-dependent vasodilatation in vitro and in vivo. We suggest that the acute scavenging of nitric oxide by heme proteins depletes the blood vessel wall of its endogenous vasodilator and permeation of heme proteins into the blood vessel wall may contribute to such sustained effects as observed in vitro.