This paper evaluates dual economy theory and research on the problem of sex differences in earnings. General Social Survey data are used to demonstrate that (1) the four most commonly used measures of the dual economy (Beck, Horan&Tolbert 1978; Bibb&Form 1977; Hodson 1977; Tolbert, Horan&Beck 1980) are extremely inconsistent when allocating women and men to industrial sectors, (2) the inconsistency results in markedly different empirical estimates of the role of the dual economy in generating sex differences in earnings, and (3) the dual economy measures provide only modest support—if any at all—for hypotheses derived from dual economy theory. In the closing section I consider the use of multiple, continuous measures of industrial structure in dual economy research.