Colombia young women (from the 1980 cohort) are more educated than men, reversing the education gap. This change has not resulted in a relative increase in the salaries of the women. The educational decision of individuals depends not only on changes in labor market conditions but also on market conditions in the marriage such marital unions and the relative size of the population by gender. Thus, this paper seeks to establish the relative importance and quantity of labor market factors and the marriage market on the change in the educational gap in Colombia. With that goal I use three quantitative approaches, the first one is the assortative matching in time, showing that in Colombia most individuals marriage with people with similar educational levels. Additionally, through a pseudo-panel by city from 1984 to 2006 shows that asymmetries in the population by gender, especially the increasing number of women reverse the educational gap. Finally, the decomposition of Oaxaca and Ransom inferred that only one of the two channels would not have achieved a reversal in the achievement gap, but both markets are required.