This paper updates the major study by Macpherson and Hirsch (1995) of the effect of the gender composition of occupations on female (and male) earnings. Using large representative national samples of employees from the Current Population Survey, cross-sectional estimates of the impact of proportion female in an occupation (or feminization) on wages are first provided, paying close attention to the role of occupational characteristics. Specification differences in the effects of feminization across alternative subsamples are examined as well as the contribution of the feminization argument to the explanation of the gender wage gap. An updated longitudinal analysis using the CPS data is also provided. This examination of two-year panels of individuals is supplemented using information from the 1979 National Longitudinal Survey of Youth which has the advantage of offering a longer panel. Analysis of the former suggests the reduction in gender composition effects observed for females in cross section with the addition of controls for occupational characteristics becomes complete after accounting for unobserved individual heterogeneity. This is not the case for the latter dataset, most likely reflecting heritage effects of discrimination in what is an aging cohort.