Abstract: In this paper we test the positive -assortative- mating (PAM) hypothesis that posits that marriages can be positively sorted along the couple’s education using Argentinian data. Using a regression approach, we find that the effect of husband’s education on wife’s education is positive and significant. Then, we empirically verify if the observed matching distribution is super-modular, i.e., the local log odds ratios of education are all positive. We conclude that the data (weakly) supports Becker’s positive assortative mating hypothesis in Argentina. The model that best describes marriage formation is more consistent with preference for own type rather than with PAM.