The relative contribution of individual-specific environmental factors varied, both across substances and racial/ethnic groups. In EA women, factors specific to individual members of a twin pair were modest but nearly twice as prominent (and statistically different) for cannabis (19%) as for cigarette use (11%). The role of E on both substances was also considerably greater in AA (42–45%) versus EA (11–19%) women. Estimation of E arises from the deviation of the MZ twin correlation from unity (i.e., rMZ≠1; (Evans et al., 2002)) and reflects person-specific factors as well as measurement error, although the latter is unlikely to be a major concern for simple binary indices of lifetime substance use in an adult population. Such differences in twin concordance across substances and ethnicities (i.e., overall: AA > EA; in EA: Cannabis > Cigarette) might be due to variations in social attitudes towards and relative availability of substances across ethnic groups, and within the EA twins, to differences in the legal status of the drugs (Boardman et al., 2010; Shanahan and Hofer 2005). For instance, while cannabis is more socially accessible in
