Cad. Saúde Públ., Rio de Janeiro, 10 (1): 17-29, Jan/Mar, 1994. A sample survey was conducted in three public ambulatory medical care units in Rio de Janeiro in order to show the demographic and nosological aspects of the clientele as well as to evaluate the quality of clinic records. The sample included 2,029 patients registered in the health clinics and 3,980 consultations from October 1, 1990 to September 30, 1991. The diagnoses were coded according to the 9th revision of the International Classification of Diseases. Females accounted for about 60% of the sample, and more than a half were between 15 and 49 years of age. Children under 10 years old made up 28% of the total. About 38% of the patients came from the poor areas of the city: the "favelas". The case files were poorly filled out, frequently lacking useful information. The most frequent cause was classified in the XVIo chapter of the ICD -illdefined causes, followed by respiratory, infectious, genitourinary and circulatory diseases. External causes predominated in one of the health clinics, which is an emergency service. There was a great demand for preventive measures, such as prenatal care and child care, even at the clinic that was traditionally devoted only to curative medicine. The most frequent specific diagnoses were upper respiratory tract infection, hypertension, gynecological problems, prenatal care, and skin diseases. There were no records for referrals to other health services. The poor quality of the medical records causes problems in terms of quality of health care. Use of such record-keeping systems, even while criticizing them, is essential to improve them. Key words: Clinic Demand; Morbidity; Quality of Records INTRODUÇÃOHá um interesse crescente no conhecimento do perfil epidemiológico para a adequação das práticas de saúde. Portanto, conhecer a demanda ambulatorial na rede pública tornou-se tarefa necessária tanto para a avaliação de serviços como para a orientação dos que trabalham com gerência, programação e planejamento em saúde (Radaelli et al., 1990).Além disso, as análises de mortalidade, já bastante difundidas, que tradicionalmente subsidiam o planejamento em saúde devem ser complementadas pela análise da morbidade, pois há um grande número de doenças e agravos à saúde que não causam o óbito, mas levam a graus variados de espoliação.Para se conhecer a morbidade na população