Theory-driven, mass media interventions prevent smoking among youth. This study examined effects of a media campaign on adolescent smoking cessation. Four matched pairs of media markets in four states were randomized to receive or not receive a 3-year television/radio campaign aimed at adolescent smoking cessation based on social cognitive theory. The authors enrolled 2,030 adolescent smokers into the cohort (n = 987 experimental; n = 1,043 comparison) and assessed them via annual telephone surveys for 3 years. Although the condition by time interaction was not significant, the proportion of adolescents smoking in the past month was significantly lower in the experimental than comparison condition at 3-year follow-up when adjusted for baseline smoking status. The media campaign did not impact targeted mediating variables. A media campaign based on social cognitive constructs produced a modest overall effect on smoking prevalence among adolescents, but the role of theory-based constructs is unclear. (Contains 4 tables.)