We relate differences in problem drinking symptoms within pairs of identical twins to their respective differences in years of schooling. Isolating in this way the influences of family background and genes, we find that an increase in schooling attainment results in a significantly lower incidence of problem drinking for men. Thus, an extra year of schooling reduces the number of health problems caused by drinking by 0.14, and the probability of developing symptoms of Alcohol Dependence (AD) by 0.06. This negative link is robust to a variety of modifications to the identifying assumptions underlying our statistical analysis. Socio-economic implications of our findings are discussed.