ABSTRACT. Objective: This study examined potential explanatory mechanisms linking childhood alcohol use onset and chronicity of adult alcohol dependence by testing the following three competing hypotheses: (1) a marker hypothesis, where early onset of alcohol use may be simply a marker for other factors that have been linked to both age at initiation and adult alcohol problems; (2) a compromised development hypothesis, where early alcohol initiation may interfere with adolescent development, which can lead to later alcohol problems; and (3) an increased substance use hypothesis, where early onset of alcohol use may lead to increased substance use in adolescence and, in turn, chronic alcohol dependence. Method: Data came from a longitudinal community sample of 808 participants recruited at age 10 in 1985. Participants were followed through age 33 in 2008 with 92% retention. Results: Childhood onset of alcohol use (before age 11), when compared with initiation during adolescence, predicted an increased chronicity of adult alcohol dependence, even after accounting for the hypothesized confounds from the marker hypothesis. In addition, adolescent compromised functioning did not mediate this relationship between early alcohol use and chronicity of adult dependence (Hypothesis 2), nor did adolescent substance use (Hypothesis 3). However, compromised functioning and substance use in adolescence predicted increased chronicity of alcohol dependence in young adulthood. Conclusions: Prevention efforts as early as the elementary grades should focus on delaying the onset of alcohol use and reducing substance use in adolescence as well as improving school functioning, reducing adolescent problem behaviors, and targeting adolescent peer networks. (J. Stud. Alcohol Drugs, 73, 379-390, 2012)