Data from a prevalence study of Organic Cerebral Syndrome and Depression in an elderly population living in three boroughs of Rio de Janeiro city are presented. The methodological issues related to interrater and test-retest reliability are discussed and the cut-off point for the instrument adapted (BOAS) established. The prevalence rates in the three boroughs were found, respectively, to be: 5.9%, 9.8% and 29.8% for Organic Cerebral Syndrome and 20.9%, 23.0% and 36.8% for Depression. The prevalence rats have been adjusted using information on sensitivity and specificity for both diagnosis. Aspects of these differences are discussed in the light of national and international literature.