by L Friberg, J Olesen HK Iversen and B Sperling, merits attention because it describes the novel application of two non-invasive techniques to address the relationship between vessel dilatation and the pain of headache.Ten migraineurs with unilateral headache were subjected to transcranial doppler sonography (to measure blood velocity within the middle cerebral artery) and single photon emission computerized tomography (to measure blood flow within the middle cerebral artery territory) in order to derive estimates of middle cerebral artery caliber changes in headache subjects. Both measurements were obtained shortly after headache onset and compared with measurements taken 30 min after administering sumatriptan. Sumatriptan abolished or significantly reduced the unilateral headache and increased middle cerebral artery velocity ipsilaterally, whereas blood flow remained unchanged. Based on a known mathematical relationship between vessel diameter and velocity when flow remains constant, the authors estimated that the middle cerebral artery diameter was increased by 20% on the headache side during the attack. The authors concluded that "headache pain was due to, or at least closely associated with, intracranial large artery dilation", that "migraine headache originates from the dilated large arteries" and that "sumatriptan... predominantly acts on pathologically distended arteries".The combined use of transcranial Doppler and xenon blood flow measurements was reported initially by Dahl and colleagues in cluster headache patients (1). In their studies, headache and vasodilatation were not tightly coupled. Bilateral, not unilateral, increases in middle cerebral artery diameter were observed during spontaneous attacks. Moreover, decreases and not increases in vessel diameter were measured at the onset of pain when nitroglycerin was used to induce painful attacks. Discounting the important possibilities that (i) the middle cerebral artery may not be the source of headache pain in cluster or migraine, (ii) the middle cerebral artery may not be representative of the responses of other large pial vessels, and (iii) the headaches of cluster and migraine are different (albeit, they both respond to treatment with 5-HT 1D agonists), the data from the Dahl study suggest that the relation between vasodilatation and headache cannot be defined simply by observing changes in vessel caliber.Many investigators, including Harold Wolff (2) have written about the association between dilatation and the pain of headache. As noted above, some have suggested that dilatation and pain are causally related (3, 4), and certainly this is true during angioplasty. However, dilatation need not be the cause of pain in this condition and a correlation does not by itself establish a cause-and-effect relationship. Dilatation can develop as a consequence of perivascular pain fiber stimulation (5). Sensory fibers projecting to meningeal arteries, veins and sinuses from the ipsilateral trigeminal ganglion (see (5, 6) for review) contain potent vasodila...