Recent investigations on multiple sclerosis (MS) distribution in Italy, based on relatively small population groups, contradict the current thesis, derived from studies undertaken in populations exceeding 300,000, of a lower spreading of the disease in Mediterranean Europe. Intensive surveys on small populations strengthen the suggestion that Italy is a high-risk area, although MS studies based on prevalence rather than incidence could also be biased by geographic variables making prevalence unsuitable for comparing MS prevalence rates found in different years and/or zones. We report a further epidemiologic study in Barbagia, Sardinia, insular Italy, undertaken both to establish the MS incidence in the last 20 years, and to estimate a new prevalence rate. Based on 31 probable incident cases, the mean incidence per year for the years 1961-1980 was 2.9 per 100,000 (3.2 if age- and sex-standardized to the Italian population). On October 24, 1981, the prevalence per 100,000 was 65.3 (77.9 if standardized to the Italian population). These results confirm that in Barbagia, MS occurs more frequently than expected in a Mediterranean area, and give further support to the idea that MS frequency in Italy is similar to that established for most central and northern European countries.