Objective A subgroup of syncope patients report migraine headaches immediately preceding or following syncope, and some respond to anti-migrainous prophylactic agents. This study aimed to describe the frequency of migrainous features concurrent with episodes of syncope and to propose clinical criteria for assessing whether a migrainous mechanism might underlie syncope. Methods This retrospective, questionnaire-based study developed criteria for syncopal migraine based on the International Classification of Headache Disorders II (ICHD-II) migraine criteria. Two hundred and forty-eight recurrent syncope subjects (>3 episodes) were stratified based on the presence ( N  = 127) or absence ( N  = 121) of a headache concurrent with syncopal episodes. Syncopal headaches were classified as either syncopal migraine (meeting ICHD-II criteria for migraine or probable migraine, without aura) or nonspecific (not meeting the criteria for syncopal migraine). The syncope groups were then compared to 199 subjects with migraine headaches using chi-square and Cochran-Armitage test for trend. Results Nearly one-third of recurrent syncope subjects met criteria for syncopal migraine. This group resembled the migraine headache population more than the syncope population in age, gender, autonomic testing, and comorbid conditions. The syncopal migraine group also reported a longer duration of syncope and a longer recovery time to normal. Finally, anti-migrainous medications reduced syncope in half of the syncopal migraine subjects. Interpretation Syncope may have a migrainous basis more commonly than previously suspected, and we suggested criteria to identify these patients. Syncopal migraine appears epidemiologically more closely related to migraine than to reflex syncope.