The aim of the present study was to compare the ability of different doses of isosorbide-5-mononitrate (5-ISMN) to cause dilatation of medium sized and small arteries, and to examine the intensity and duration of any headache produced. Ten healthy volunteers each received 3 doses of 5-ISMN and placebo on separate days. The diameters of the radial and superficial temporal arteries were repeatedly measured with high frequency ultrasound and pain was scored using a 10 point verbal scale. A clear dose-relationship was found for plasma concentrations and headache, and for changes in the diameter of the temporal artery, but not for the radial artery. It is concluded that headache after 5-ISMN is caused by arterial dilatation or by mechanisms responsible for the arterial dilatation. Ultrasound monitoring of arterial diameters is an important and sensitive tool in the evaluation of nitrates and other vasodilators.