This paper described a number of critical assumptions and methodological issues that influence the interpretation of data from family, twin and adoption studies of substance abuse. Also, using data from a recent study, it examined the role of one methodological issue (method of zygosity determination) and one critical assumption (representativeness of twins as research subjects) to determine the validity of data in twin studies of substance abuse. The accuracy of the questionnaire methods for zygosity determination in alcoholic twins was similar to that previously reported for nonalcoholic twins. Also, alcoholic twins were found to be representative of both nonalcoholic twins and alcoholics in general.