To assess the prevalence of depression and associated factors in elderly people in Saudi Arabia. A cross-sectional national survey of the elderly population of Saudi Arabia was conducted between January 1994 and December 1995. A stratified two-stage sampling technique was used to select the study subjects. In the first stage, a random sample of primary health care centres (PHCs) was selected in each of the five administrative regions of the country. The family health records (FHR) of each selected PHC were reviewed and a list of the elderly (60 years and over) was compiled. In the second stage, a sample of the subjects was selected from the FHR and contacted for a personal interview. The subjects' physical, social and environmental health status were assessed by an interview during which a structured questionnaire was completed. The Geriatric Depression Scale (GDS) was used to screen for depression. Other items of information in the questionnaire included subjects' sociodemographic characteristics, activities of daily living (ADL), laboratory and radiological investigations. The total number of elderly subjects included in this study was 7970. Their mean age+/-standard deviations was 68.8+/-7.7 (male 69.1+/-.7.7; and female 67.7+/-7.5) years. Depressive symptoms were reported in 3110 (39%) of the subjects, and 8.4% were in the severe depressive symptoms score group. Personal characteristics that correlated strongly with depression were poor education (p=0.001), unemployment (p=0.001), divorced or widowed status (p=0.001), old age and being a female (p=0.001). Living in a remote rural area with poor housing arrangements and limited accessibility within the house and poor interior conditions were also significantly associated with high depressive symptoms (p=0.001). Limited privacy, such as having a particular room specified for the elderly, was associated with more depressive symptoms than sharing a room with another person (p=0. 001). Lower incomes inadequate for personal needs as well as depending on charity or other relatives were associated with more cases of depression (p=0.001). The proportion of cases of depression correlated positively with the number of medical diagnoses and medications received (p=0.001). Significant depression was associated with loss of a close relative, living alone and limited participation in recreational activities. Perception of poor health and dependence on others for daily activities were associated with more depressive symptoms (p=0.001). Also health problems, especially faecal or urinary incontinence, were associated with more depressive symptoms (p=0.001). Depressive symptoms are common among Saudi elderly. Detection and management rates were low. Primary care teams could help these patients if properly trained. A simple instrument such as the Geriatric Depression Scale is useful and easily administered.