OBJECTIVE. The study examined differences in the use of mental health services by persons living in different types of households. METHODS. Data were obtained from the Epidemiologic Catchment Area (ECA) survey, a multistage probability sampling procedure of households in five U.S cities conducted in 1981-1984. Data from 12,417 persons living in four of the five ECA sites (Baltimore, St. Louis, Durham, North Carolina, and Los Angeles) were classified according to household categories used by the U.S. Bureau of the Census. Logistic regression analysis was used to examine the relationship of mental health service use to household type, relationship of survey respondent to head of household, household income and size, number of children in the household, ECA site, presence of mental disorders, and sociodemographic factors. RESULTS. Use of mental health services varied by type of household but not by individual within the same household. Persons in female-headed families and persons living alone had the highest use of mental health services, even after other variables were controlled. CONCLUSIONS. The findings highlight the need for outreach efforts to determine needs and preferences for health, mental health, and social services by persons in various households, especially persons in female-headed families and persons living alone.