Investigations over more than four decades have established that multiple sclerosis (MS) has an unequal geographic distribution [36]. Such studies, while of some intrinsic interest, are valuable chiefly because of the light they may shed upon possible etiologies, risk factors, and the pathogenesis of the disease. Before examining the fruit of almost a half century of effort in this field, it is necessary to remind ourselves of the technical problems which had to be overcome in order that data on MS distribution would be meaningful. One such problem is diagnosis. As in any scientific study, we must be sure that the entity we are looking at is relatively "pure". For MS, the difficulties in designating the affected have been (and continue to be) the lack of an objective, pathognomonic test to support the clinical impression. The problem has been resolved (though certainly not solved) by using clinical criteria to define the individual case. These criteria include occurrence of remissions and exacerbations (time scattering) as well as evidence on examination or documented evidence of multiple lesions of central white matter (place scattering) [41]. Various investigators have modified these criteria somewhat to include an age-at-onset limitation [52], and a few recommend including elevation of gamma globulin [56] in the cerebrospinal fluid or oligoclonal banding [58]. However, it is reassuring, at least for the present, that experienced neurologists have a very high rate of agreement as to the "purity" of the entity for patients who fulfill the most stringent of these clinical diagnostic criteria. Of course, discovery of an objective test for MS may radically affect current concepts of MS distribution.Techniques of case ascertainment have evolved over the last several decades, but for the most part, data on MS distribution based on mortality data [35] have been supported by the more sophisticated morbidity data, which use living patients with MS to measure prevalence and incidence rates [24]. The morbidity data (Fig. 1) suggest a gradient of MS frequency strongly correlated with geographic latitude. The higher the latitude both north and south of the equator, the higher the apparent frequency of MS. Almost everyone agrees that this distribution is not a function of the availability of medical care facilities, though such facilities obviously influence the number of cases identified in a given area [59]. Therefore, one should not look to the absolute number of cases identified, but rather to whether the frequency in a given area is low, intermediate, or high. The cut-off limits for each of these areas is arbitrary, but most investigators agree that an area with less than ten living cases per 100,000 population at anyone time constitutes a low prevalence area; one with 10-35 cases, an intermediate area; and one with more than 35 cases per 100,000 population, a high MS prevalence area. Some areas, such as the Shetland and Orkney islands [49] and Denmark [37], allegedly have up to a hundred or more