Several rigorously evaluated Early Childhood Interventions (ECIs) have been proven effective in improving the long-run outcomes of their beneficiaries. However, a common critique is that these interventions, which are usually aimed at improving children’s intelligence, are not able to increase IQ scores in the long-run. In many cases, they fadeout: initial positive effects diminish or disappear some years after children have left the program. This fact is a motivation for the objective of this proposal, which is to understand why some impacts on skills from ECIs fadeout while others persist. I will use data from 7 of the most important experimental ECI’s with long-term follow- ups that exist (Perry, Abecedarian, Head Start Impact Study, Early Head Start, Early Training Project, Project CARE and Infant Health and Development Project). I propose an integrated approach to analyze impacts on three groups of measures: (i) intelligence quotient (IQ), (ii) academic achievement and (iii) personality (social-emotional) skills. For the programs that present impacts, I will examine the persistence of that initial effect on all of those skills, over time. Moreover, I propose to analyze the determinants of the impacts of the program, and quantify the contributions of the different factors. Preliminary findings on IQ show that fadeout is widespread across ECIs. This is true even for programs with important adult benefits in labor market outcomes. After taking into account the natural stability of IQ and other factors, I find that unexplained fadeout is centered around school entry age. This finding can be helpful for understanding the nature of the fadeout phenomenon, and ways to ameliorate it.