The operational and diagnostic criteria for migraine and all other headache disorders released in 1988 by the International Headache Society are universally considered reliable and exhaustive. These criteria, however, cannot be considered as satisfactory for population-based studies on migraine prevalence, especially if adolescents are the subjects of the study. Using these diagnostic criteria, we conducted an epidemiological study in order to assess the prevalence of migraine headache in a student population aged 11 to 14 years. Our survey made it possible to code IHS 1.1 (migraine without aura) in 2.35%, IHS 1.2 (migraine with aura) in 0.62%, IHS 1.7 (migrainous disorders not fulfilling migraine criteria) in 1.52%, and IHS 13 (headache not classifiable) in 1.38% of the examined pupils. In adolescents, the low prevalence estimates of migraine headache coded IHS 1.1 and the relatively high prevalence estimates of headaches coded IHS 1.7 and IHS 13 have appeared to be a consequence of the rigidity of some operational diagnostic criteria of the recent IHS classification rather than of the geographical, environmental, or socioeconomical peculiarities of the cohort. Therefore, in order to improve the reliability and the exhaustiveness of the IHS classification by increasing its sensitivity, we believe that minor modifications of the diagnostic criteria are necessary. Within these revised criteria, the subitem "moderate or severe intensity" of pain headache should become mandatory, whereas the lower limit of the criterion "duration of pain" should be reduced to 1 hour.