The sampling methodology used in an evaluation of the standards of living and health in the urban population of Botucatu, SP, Brazil is discribed. Data were obtained through domiciliar interviews and the accuracy of sampling was evaluated for each different stratum and in accord with age and sex. External criteria for socio-economic stratification of the population such as family or "per capita" income were compared. It was found difficult to work with the concept of social classes when using socio-spacial zoning in domiciliar interviews or epidemiological studies based on the normal registers.
Health surveys; Sampling studies; Socioeconomic factors