Agreement among reporters on features of family life, whether family members or outside observers, is considered to be low. This study, which involved a national sample of 720 families comprised of identical and fraternal twins, full siblings, half siblings, and biologically unrelated stepsiblings, examined the issue of low interrater agreement by decomposing the common and unique variance among parent, child, and observer reports of parenting behaviors (warmth and negativity) into genetic and environmental factors. Quantitative genetic analyses were employed to decompose the “Social” level of perception (common variance among parents, children, observers), the level of “Family” subculture (common variance only among parents and children), and the unique “Individual” level into genetic and environmental components. It was predicted that genetic factors would account for substantial portions of the variance at the Social and Family levels; nonshared environmental factors were expected to influence variance unique to child reports; and shared environmental factors were expected to influence variance unique to parent reports. A second and related aim of the study was to examine the subjective – objective dimension of genetic effects on measures of the environment. Results of model-fitting analyses generally supported the predictions for parental warmth and negativity at the Family and Individual levels. At the Social level, genetic factors were predominant for parental negativity and shared environmental factors for parental warmth. The findings are discussed in terms of genetically influenced child effects on parenting and methodological difficulties in constructing latent variables.