The main goal of this paper is to analyse the impact of child marriage and adolescent pregnancy in the education of girls in the most populated countries of Latin America and with similar economic and demographic characteristics, Brazil and Mexico. The methods used are descriptive statistics and logistic regression, analysing cohorts and periods. The data used for Brazil are from the National Demographic and Health Survey of Children and Women (PNDS-2006), National Health Survey (PNS-2013), National Household Sample Survey (PNAD-2004,2014), and for Mexico the Mexican Family Life Survey (MxFLS), 2005-2006, 2009-2012. Both countries struggle with school dropout related to social and health inequalities, as many young people drop school to work, for lack of interest and for early pregnancy and union. The preliminary results shows that teenage-specific fertility rates dropped over the decade, however 1 in every 4 pregnancies are from adolescents. The study identifies that girls from 10-14 years remains the most vulnerable group. Marriage and informal union increases the probability of pregnancy but are not necessary conditions for a significant share of the girls. For the majority of young mothers the pregnancy was not planned and, in Brazil, abortion still afflicts 12% of them. There are evidences that the postponement of union and pregnancy increases the probability of girls increasing their education level, with stronger effect for girls in socioeconomic disadvantage and self-identified as black or indigenous. The analysis takes in consideration the legislation and current policy regarding girl's reproductive health and compare the results with the interventions and past trends in United Kingdom whith the highest teenage fertility rate in Western Europe, although with fast declining rates.