Experiments were designed to compare the relaxant activity of the new nitrate ester ITF 296 in isolated arteries and veins of the dog and to determine the extent of modulation by the endothelium of the response to the compound. Rings (with and without endothelium) of coronary, basilar, mesenteric, and femoral arteries and of mesenteric, femoral, and saphenous veins were suspended in organ chambers for measurement of changes in isometric tension. In all blood vessels without endothelium, ITF 296 caused concentration-dependent relaxation. The order of potency (from measurements of ED50) was: basilar artery > or = coronary artery > femoral artery > mesenteric vein > femoral vein > saphenous vein > mesenteric artery. The maximal relaxation to ITF 296 was greater in the arteries (with the exception of the mesenteric) than in veins. In all blood vessels except for the basilar artery, nitroglycerin caused larger relaxations than ITF 296: The ED50 for nitroglycerin was comparable in all blood vessels studied except for the mesenteric artery, where it was less. ITF 296 did not cause hyperpolarization of vascular smooth muscle in coronary arteries either with or without endothelium. The presence of endothelial cells blunted the relaxation to ITF 296 in basilar and coronary arteries, but not in the other blood vessels studied. These results demonstrate that ITF 296 has a direct inhibitory effect on vascular smooth muscle, which is most pronounced in coronary and cerebral arteries. These findings are in line with the vascular selectivity profile reported for the compound in vivo.