This study compared two major Federally sponsored surveys of adolescent substance use and assessed the impact that methodological differences have on the prevalence estimates they generate. The Monitoring the Future Survey, a school-based survey, was compared to the National Household Survey on Drug Abuse, a household survey of the population aged 12 years and older. Response rates were higher in the household survey due to high rates of refusal in the school based survey. The school survey has a larger overall sample size, but sampling errors more similar than one might expect, because of the larger design effects in the school survey. Rates of drug use obtained were larger in the school survey than in the household survey, possibly because of greater under-reporting in the household setting than in the classroom and the different questionnaires used in the two surveys.