Previous attempts to reduce the effects of television advertising on children's purchase requests have had little success. Therefore, we tested the effects of a classroom intervention to reduce television, videotape, and video game use on children's toy purchase requests, in a school-based randomized controlled trial. Third- and fourth-grade children (mean age, 8.9 years) in two sociodemographically and scholastically matched public elementary schools were eligible to participate. Children in one randomly selected elementary school received an 18-lesson, 6-month classroom curriculum to reduce television, videotape, and video game use. In both schools, in September (before intervention) and April (after intervention) of a single school year, children and parents reported children's prior week's purchase requests for toys seen on television. After intervention, children in the intervention school were significantly less likely to report toy purchase requests than children in the control school, with adjusting for baseline purchase requests, gender, and age (odds ratio, 0.29; 95% confidence interval, 0.12-0.69). Among intervention school children, reductions in self-reported purchase requests were also associated with reductions in television viewing. There was no significant difference between schools in parent reports of children's requests for toy purchases. These findings suggest that reducing television viewing is a promising approach to reducing the influences of advertising on children's behavior.