Early age-of-onset drinking is associated with alcohol problems and related, high-risk behaviors. We analyzed data from 18- to 34-year-old respondents from the 2001 to 2002 National Epidemiological Survey on Alcohol and Related Conditions (NESARC) to determine to what extent early age-of-onset drinking increased risks for prescription drug misuse (PDM), defined as 1 or more positive responses to: "ever having misused sedatives, tranquilizers, painkillers or stimulants, obtained either as prescriptions or from indirect sources." Lifetime prevalence of PDM was 15.4% among 8306 "drinkers" and 3% among 4652 "nondrinkers." Unadjusted odds for PDM for both men and women increased with each successively younger drinking age-of-onset, reaching a 10-fold risk at <14 years (odds ratio [OR], 10.7; 95% confidence interval [CI], 7.43-15.3) for men and women combined). In adjusted analyses, early age-of-onset marijuana use among drinkers reduced these odds and independently increased the risks for PDM 2-fold (adjusted OR, 2.07; 95% CI, 1.17-3.64). Lifetime alcohol dependence independently predicted PDM (adjusted OR, 2.43; 95% CI, 2-2.96) and obscured the association of early drinking, but not of early marijuana use, with PDM. This finding suggests a specific mediating effect of alcohol dependence between early drinking and PDM. Findings support the need for effective programs to prevent and reduce harm from early-onset drinking and from the associated risk of alcohol dependence and prescription drug misuse.