Alcohol use usually starts in early adolescence. While the greater proportion of young people develop adaptive patterns of drinking, many drink at harmful levels and may be at risk for future alcohol-related problems. Findings from the empirical literature suggest that universal prevention programs may delay onset of drinking among low-risk baseline abstainers; however, there is little evidence supporting their utility for at-risk adolescents. Further research is needed on how risk and protective factors interact to determine substance use trajectory, and intervention outcomes that take substance use trajectories into account may capture change more effectively than the use of absolute measures of substance use. Indicated prevention programs may benefit from modulations that account for adolescent individuation and identity formation. It is argued that motivational interviewing within a harm reduction framework is well suited to adolescents.