We seek to add to the literature on the causes of the motherhood-wage penalty and to the literature on the contribution of industry to the gender wage gap by using a longitudinal history of industry and firm size instead of a point-in-time measure.  Because we observe firm characteristics at multiple points in time, we can observe the gender distribution across industries when workers are just beginning their careers, after children are born, and in later, post-fertility years.  Initially, we consider all women prior to childbirth and compare their distribution across industry and firm size to that of men of the same age.  We then ask how this distribution changes as women have children (or not) and how it further evolves as the children age.  This allows us first to determine whether mothers initially switch to industries and firms likely to be “family-friendly” and whether this results in a wage cut.  Second, we can observe whether they switch back again later in life and return to similar wage levels as their childless female peers or whether these changes early in the career continue to impact women after the childbearing years have passed.  It also allows us to ask whether childless women are initially different from men but adjust their industry and firm-size choices over time so that their earnings profiles are more similar to men’s, and if so, what impact this has on the male-female wage gap.  We use a traditional decomposition method that accounts for differences in the gender ratio and wages in each industry and firm-size category as well as for observed characteristics and unobserved heterogeneity of people and firms. Decomposition enables us to parse the gender and motherhood wage gaps into components that are attributable to choices women make about industry and firm size.