Adoption studies cleverly leverage a naturally occurring disjoining of parents’ genetic and environmental influences to make causal inferences regarding their effects. Unfortunately, extant studies are typically restricted to examinations of rank-order (dis)similarity between reared-apart and reared-together relatives, ignoring the well-documented presence of mean differences in outcomes across family members. We sought to fill this gap, jointly examining both mean differences and indices of similarity in 424 linked adoptive families participating in the Early Growth and Development Study (EGDS) using multilevel models. Results indicated that although adopted children’s achievement was modestly correlated with that of their reared-apart biological relatives, they scored 3.01 to 15.09 points (0.20 to 1.00 standard deviations) higher than those relatives. What’s more, these differences were partially attributable to rearing-home literary opportunities. We conclude that, by omitting considerations of mean differences between reared-apart relatives, behavioral geneticists may have inadvertently failed to detect and consider key environmental effects.