Based on a hypothesis that the cause of Alzheimer's Disease (AD) is multifactorial, we explored the possible links between the geographical distribution of 102 subjects diagnosed as “definite,” “probable” or “possible” cases at the onset of the disease and the social ecology of the Chicoutimi-Jonquière (CJ) urban area in the Saguenay region of Québec. We were also interested in the level of education of the 74 “definite” and “probable” cases. The use of factor analysis (principal components) and cluster analysis (Ward method) made it possible for us to identify socioeconomic zones and homogeneous social areas in CJ. The first results indicate that on the one hand, the male/female ratio of the cases differs statistically from that of the at-risk age group in the general population. The disease affects more women than men. On the other hand, there is significant social differentiation in CJ. Nevertheless, the 102 subjects are randomly distributed among the three socio-economic zones as well as among the eight social areas. Analysis of the geographical distribution of “definite” and “probable” cases also suggests a random distribution. The random nature of the distribution of subjects is confirmed by a very strong correspondence between the distribution of cases observed when counting them by enumeration area or census tract and Poisson's theoretical distribution. There was a significant difference between the subjects' level of education and that of the 65 years old and over in the reference population. A more in-depth case-control study might shed more light on this aspect.