Determining the rate to which various psychiatric illnesses are familial is one widely used method for validating diagnostic categories and determining the likelihood of genetic or nongenetic patterns of transmission. Data for these studies can be collected through direct interview of all available relatives (the family study method) or by obtaining information indirectly from the patient and other family members (the family history method). Information based on direct interview is usually considered to be more accurate, although the family history method permits collection of data on a larger and more comprehensive group of relatives. We explored the extent to which data collected by these two methods were in agreement. In general, the results confirmed the usefulness of the family history method. Although it has some limitations, such as underreporting, it has respectable sensitivity for many major diagnoses when broad but well-specified criteria are used.