The effects of trade on women vary by socio-economic characteristics, sector and country. This paper assesses how well such effects can be captured by a gendered social accounting matrix (SAM) and computable general equilibrium (CGE) model. These are applied comparatively to Bangladesh and Zambia to highlight how differences in resource endowments, labor market characteristics and socio-cultural norms shape the way in which trade expansion affects gender inequalities. The paper also compares simulation results to other approaches in the gender-and-economics literature, discusses strengths and limitations of the CGE methodology, and provides suggestions for further research.