For general practitioners and in epidemiological studies, migraine poses diagnostic problems. It is a polymorphous disease, with diagnosis based on a subjective description of symptoms. The International Headache Society (IHS) proposed a set of rigorous criteria for the diagnosis of migraine. We studied its validity on 171 outpatients of headache clinics and on 96 headache sufferers detected during an annual checkup for workers. The criteria showed an excellent specificity, however sensitivity in both samples was < 50%. This lack of validity could be due to IHS criteria being too restrictive. We therefore elaborated and validated our own diagnostic tool, based on the IHS criteria, to a sensitivity of 95% and a specificity of 78%. We propose our tool for epidemiological studies.