A conviction for drug possession blocks some of the most common pathways through which individuals from low income families achieve upward economic mobility in the United States, such as access to higher education, entry-level employment, and military service. These considerations are of growing importance because the number of drug-related arrests have nearly quadrupled since 1980. This article estimates the effect of a conviction for drug possession on earnings mobility using a sample of individuals born between 1980 and 1984, some of the first cohorts to come of age in the context of intensive U.S. drug criminalization and enforcement. To distinguish the effect of a drug conviction from the effect of drug use or general criminality, 1 compare mobility among individuals with drug convictions to control groups who self-report significant drug use and who have had interactions with the criminal justice system that did not lead to a drug conviction. I find that relative to these groups, a drug conviction reduces the probability of transitioning upward from various points in the lower half of the income distribution by 10-15 percentage points, or as much as 50%, and that these effects are substantially stronger for non-whites than for whites. These findings suggest that a policy of decriminalizing nonviolent drug possession would substantially increase intergenerational mobility among low income populations, and this effect should be weighed alongside more conventional costs and benefits in formulating optimal drug policy. (JEL J38, J15, K42) I. INTRODUCTION Major earnings gains from one generation to the next are a relative rarity in the United States, especially at the lower end of the income distribution. Recent definitive estimates from Chetty et al. (2014) using administrative tax records indicate that individuals born to parents in the bottom quintile of the income distribution have a 33.7% chance of staying in the bottom quintile of the income distribution themselves, a 71.9% chance of staying in the bottom half, and only a 7.5% chance of attaining the top quintile. (1) A large theoretical and empirical literature has investigated possible mechanisms through which parents transmit economic status to their children, such as the genetic inheritance of traits associated with earnings and parental investments in their children's human capital (see Black and Devereux 2010 and Solon 1999 for reviews). However, even taking these biological and family-level transmission processes as fixed, both common sense and existing research suggest a few basic pathways through which individuals coming from the lower portions of the income distribution effectively achieve improved earnings. In particular, children of low income parents who realize improved economic status frequently do so by increasing their educational attainment, particularly at the postsecondary level by obtaining entry-level employment that offers human capital acquisition and internal promotion opportunities or by serving in the military, which often includes on the job training as well as access to educational financing. This study is motivated by the observation that these basic strategies for realizing upward mobility are severely hampered by a conviction for drug use. Specifically, in the United States, individuals with a drug conviction are typically ineligible for federal student aid, introducing a potentially important credit constraint for higher education. Criminal convictions have also been shown to disrupt early life human capital accumulation more generally (Aizer and Doyle 2013; Azarnert 2010), and a drug conviction makes it difficult or impossible for an otherwise qualified individual to enlist in the military. Large literatures additionally find that criminal convictions directly reduce entry-level employment opportunities, where a clean criminal record is often a primary qualification (Grogger 1995; Kling 2006; Waldfogel 1994a, 1994b; Western 2002). …