Head Start is a federally funded preschool program for poor children designed to help close the gap between those children and their more advantaged peers before they begin public schooling. Given that Head Start appears to have positive long-run impacts on its recipients, a natural and important next question to ask is: What are the channels through which Head Start improves these long-run outcomes? In other words, what skills, behaviors, attitudes or circumstances does Head Start influence that result in these long term benefits? One possible explanation is increased cognitive ability. However, researchers have shown that cognitive test score gains due to Head Start participation fade away well before such long-term benefits are realized (Currie and Thomas, 1995; Deming 2009). Another possible explanation is that long run impacts are due to improvements in "non-cognitive skills" such as self-esteem, social skills and motivation. Given that Head Start curriculum aims to improve the "whole child" and families receive a host of health and social services, this is a reasonable but unconfirmed hypothesis. The purpose of this project is to determine the extent to which cultivating non-cognitive skills is a mechanism through which Head Start successfully improves long-term outcomes. Results show that Head Start participation has positive impacts on particular non-cognitive traits of children, but has little if any detectable impact on the traits of adolescents. These results are driven exclusively by gains by female Head Start participants, which is consistent with past research that early childhood interventions have more lasting effects on girls than boys (Anderson, 2008) and that girls in general have higher non-cognitive skills than boys (Jacob, 2002). The results show suggestive but inconclusive evidence that these non-cognitive skills improve the long-term outcomes that Head Start has been shown to impact. (Contains 5 tables.)