Data for this analysis came from a cross-sectional study on dementia, depression, and disability conducted in Zurich and Geneva in 1995/96. The random sample stratified by age and gender consisted of 921 subjects aged 65 and more. Based on the Canberra Interview for the Elderly, depression was assessed by means of psychogeriatric assessment scales (PAS) according to DSM-III-R criteria. The number of depressive symptoms (NDS) and the prevalence rate of depression (PRD) were computed for the whole sample as well as according to age and gender. To evaluate the independent effects of age as well as gender with regard to the risk of being depressed, multivariate analyses were conducted. On average, 13% of females vs 8% of males reported having at least one depressive symptom. The PAS yielded 298 (41.8%) subjects without depressive symptoms, 341 (50.2%) with 1-3 symptoms, and 60 (8.0%) with four or more. The average NDS was 1.27 (95% CI 1.16-1.39). For females, NDS values statistically significantly higher than those for males were calculated (1.53, 95% CI 1.35-1.70 vs 1.05, 95% CI 0.90-1.20). The NDS increased significantly with age. Subjects with low education levels and being divorced or widowed had statistically significantly higher NDS values than highly educated, married, or single persons. There were strong positive associations between NDS, dementia, and activities of daily living. Multivariate regression analysis revealed gender - however, not age - as a strong risk factor for NDS. Overall PRD amounted to 8.0% (95% CI 5.7-10.2%). Females had statistically significantly higher PRD values than males (10.4%, 95% CI 7.0-13.9% vs 3.9, 95% CI 2.0-5.9%). The PRD increased substantially with age. After adjustment for other risk factors, multivariate logistic regression analysis confirmed the positive statistically significant association between age, gender, and depression.