Four potential biases incurred by using a sample of wives of male respondents as a sample of women for sex comparisons of the occupational attainment process were examined and empirically assessed. Using data from the Wisconsin longitudinal study of high school seniors, the extent and nature of two potential biases were considered: (1) a sampling bias due to the exclusion of not currently married women, and (2) the bias arising from the omission of data in crucial variables for the sample of wives. Results indicate that the omission of the former appears to have minimal (if any) effect on sex comparisons of occupations attainment, while the absence of the latter (an expanded list of variables for wives) appears to have more deleterious effects on research outcomes. In particular, the lack of availability of measures of socioeconomic index (SEI) of first job for wives affects the sex comparisons. Using the expanded list of variables also indicates differences in the mechanism by which education affects occupational attainment. (MEK)