In the US labor market, married men and women earn higher wages than their single counterparts. At the same time, individuals with higher cognitive and non-cognitive skills are more likely to be married. We extend the frameworks of Altonji and Pierret (2001) and Pinkston (2009) to the case of marriage and find no evidence that employers use marriage to statistically discriminate workers. Contrary to what statistical discrimination implies, the returns to being married increase with labor market experience. For women without experience being married is associated with a penalty. However, as experience increases, the relationship between wages and being married becomes positive. These findings are valuable in building a better understanding of the determinants of the marriage wage premium.