Domestic violence is a crime that affects women across the world. Tanzania is no exception. However, only a few studies have examined the relationship between domestic violence and women’s employment status in Tanzania. Moreover, such studies have methodological limitations. Using a 2-stage linear probability model on a nationally representative sample, this study shows that the impact of women’s employment status on reducing domestic violence is much greater if endogeneity is controlled for. We conclude by arguing that there is a need for policies that enhance women’s employability and we call for greater rigour in research analysis to avoid drawing misleading conclusions.