We investigate experimentally the impact of quota policies on gender discrimination in hiring decisions by testing whether affirmative action increases female employment. We also ask whether firm performance is affected by such policies. Our experiment consists of three treatments. In the baseline (no quota) treatment, groups of two employers and six potential job candidates are formed. Employers have to hire two workers based on information on candidate characteristics including gender and years and subject of study. The second, low penalty, treatment is identical to the baseline except there is a quota such that at least half of the employees hired must be women. If this quota is not respected, the firm has to pay a penalty. The last, high penalty, treatment is the same as the low penalty treatment except that the penalty is significantly higher. We find that women are ranked unfavorably in the absence of a quota, and the introduction of a quota significantly reduces gender discrimination. Firm performance is not affected by the introduction of quotas.