This article explores the relationship between occupational segregation and the wage differences between men and women using the Fluckiger & Silber (1999) decomposition. This methodology is an adaptation of the traditional Oaxaca (1973)-Blinder (1973) decomposition that allows separating the sex wage gap into three additive components: 1. The part of the gap explained by differences in productivity between sexes. 2. The fraction associated with differences in returns to productive characteristics and 3. A part related to differences in group size –segregation-. Despite the statistical difference in productive characteristics and returns across occupational groups, we find that the segregation component of the decomposition is extremely sensitive to the way groups are defined. Since occupational groups are defined ad hoc, further insight is needed in order to determine how important segregation really is to explain wage differences.