Aims To examine the relationship between exposure to different forms of alcohol advertising and subsequent drinking among US adolescents and assess whether exposure to an alcohol and drug prevention program mitigates any such relationship. Design Regression models with multiple control variables examined the relationship between exposure to alcohol advertising in grade 8 and grade 9 drinking for two groups of South Dakotan adolescents: (1) seventh-grade non-drinkers (n = 1206) and (2) seventh-grade drinkers (n = 1905). Interactions between the intervention program and the significant advertising predictors were tested. Setting Forty-one middle schools in South Dakota, USA. Participants A total of 3111 seventh-graders followed through grade 9. Measurements Advertising variables were constructed for four types of alcohol advertising—television, in-store displays, magazines and concession stands. Other predictors tested included measures tapping social influences, social bonds, problem behavior, alcohol beliefs, television exposure and demographics. Findings For seventh-grade non-drinkers, exposure to in-store beer displays predicted drinking onset by grade 9; for seventh-grade drinkers, exposure to magazines with alcohol advertisements and to beer concession stands at sports or music events predicted frequency of grade 9 drinking. Although exposure to television beer advertising had a significant bivariate relationship with alcohol use for grade 7 non-drinkers, it was not a significant predictor of drinking for either group in multivariate analyses. Participation in the prevention program, ALERT Plus, reduced future drinking for both groups and counteracted the effect of in-store beer displays. Conclusions Several forms of alcohol advertising predict adolescent drinking; which sources dominate depends on the child's prior experience with alcohol. Alcohol prevention programs and policies should help children counter alcohol advertising from multiple sources and limit exposure to these sources.