In this research we examined race, sex, and age differences in the factorial structure of Liang's (1985) model of subjective well-being that integrates the Affect Balance Scale and the Life Satisfaction Index A. In particular, we viewed the covariance structure of the items as a function of several parameter matrices. We analyzed the factorial invariance by testing hypotheses involving the equivalence constraints of one or more parameter matrices with regard to the following: White and Black subsamples, men and women, and the young-old and the old-old. Data for the research came from the 1974 Harris National Council on Aging Survey, Myths and Reality of Aging in America. Analysis of covariance structures, or LISREL, was used to assess the factorial invariance. Replicated race differences were found in the factorial structure, but sex and age differences were not found. Consistent race differences were found for the second-order factor loadings for negative affect.