One major challenge for intervention regarding alcohol is to target children with age-appropriate strategies while predictive risk and protective factors are still forming. Most intervention research has focused on children of preadolescent or adolescent ages, but recent work suggests that interventions may be most effective with children prior to their first experience with making drinking decisions. A study therefore evaluated the effectiveness of a media literacy lesson on 246 third graders from a northern midwestern town. The subjects were randomly assigned to four groups. The hypothesis that critical viewing training would affect children's perceptions of alcohol norms, alcohol portrayals in advertising, and alcohol-related behavior, was tested via one-way analysis of variance (ANOVA) and multivariate analysis of variance (MANOVA). The MANOVA, which tested for within-subject change over time between pretest and posttest, was significant only for perceptions of realism. The ANOVAs, however, showed significant group differences at posttest for the treatment and nontreatment groups on their understanding of persuasive intent, perceptions of realism, perceptions of social norms for alcohol use, and behavior. These results provide support for the value of media literacy programs at the third-grade level as a way to minimize the effects of alcohol advertising on the development of alcohol expectancies and related behavior among children. (Contains 3 figures, 7 tables of data, and 45 references.) (TB)