The aim of the present study was to test the hypothesis that patients whose headaches are usuallypreceded by prodromes would derive more benefits from biofeedback than those who do not have anywarnings. Accordingly, 39 female patients were assigned to either of two conditions: classical migraine(n=17) or common migraine (n=22) depending on the characteristics of their headaches. The classicalmigraine condition included only patients whose headaches were usually or always associated with visualprodromes while the common migraine condition included patients whose headache were never or rarelyassociated with any kind of prodromes. All patients completed 5 weeks of daily self-monitoring ofheadache activity and medication intake before and after treatment, and at 6-month follow-up. Resultsshowed that patients in the classical migraine condition improved significantly more than those in thecommon migraine group. Various explanations for the differential effects of biofeedback in the treatment ofclassical and common migraine are discussed and suggestions for further research are made.