Fifty-six years after the introduction of affirmative action in employment in the U.S., there is a lack of consensus regarding the effect of this policy on workers’ careers (Holzer and Neumark, 2000). This paper contributes to fill this gap by building and analyzing a dataset that allows us to quantify the effects of affirmative action in employment on workers’ labor market outcomes. This paper circumvents prior data restrictions by constructing the first administrative database containing worker-level information (from the Longitudinal Employment Household Dynamics) as well as the federal contractor status of workers’ employers (from the Equal Employment Opportunity Commission Data and Federal Procurement Data). We estimate the causal effects of affirmative action on workers’ outcomes exploiting different features specified by the legal obligations of the regulation in a regression discontinuity setting.