A descriptive epidemiological survey was extended to the whole province of Sassari, northern Sardinia between latitudes 40 degrees 30' N and 41 degrees N. Results showed a crude total prevalence rate of 144.4 per 100000 on prevalence day (31 December 1997), and an onset-adjusted prevalence rate of 149.7 per 100000. The total average annual incidence rate was 4.9 per 100000 for the whole time interval studied (1968-1997), having increased from 2.0 in 1968-1972 to 6.8 in the last quinquennium considered. A substantial improvement in MS case ascertainment due to the introduction of new diagnostic procedures might account for such rates in Sardinia as well as in other Italian regions. However, when comparing our data with those obtained in the province of Ferrara, in the same time frames (1968-1997), a nine-fold versus a five-fold increase of MS prevalence was detected in Ferrara and Sassari, respectively. MS incidence temporal trend also notably increased in Sassari, but remained substantially stable in Ferrara. The progressive shortening of the time interval between clinical onset and diagnosis, and the proportion of benign-mild MS cases, were similar in both studies, thus suggesting a similar diagnostic accuracy. In our opinion, the repeatedly assessed increase of MS frequency in our province, at least partially does represent an actual rise of MS risk among Sardinians, thus disproving the latitude gradient-based theory (i.e. prevalence rates correlate with geographical latitudes) and supporting the hypothesis of a "Sardinian focus" of MS in a genetically susceptible population.