Sandler, Youdim and Hannington noted an inability to oxidise tyramine and phenylethylamene in dietary and non dietary migraine sufferers-l,883 female volunteers in Britain completed questionnaires detailing food intake during the 24 hours prior to spontaneous migraine attacks. Analysis of 2,313 attacks are reported. This revealed consumption of chocolate in 33%, cheese 40%, citrus fruits 21% and alcohol 23%. Fasting (5 hours day or 13 hours overnight) occurred in 67%. No dietary factors were isolated in 5%. Only 14% thought food caused their attacks and 2% blamed fasting. Sensitising factors analysed included age, time in menstrual cycle, pill taking and hysterectomy. The findings suggest that in double blind trials designed to elucidate dietary factors, one needs to consider the patient's intake during the 24 hours prior to the administration of the test capsule. A new approach to treatment is suggested with elimination of specific foods and avoidance of fasting.