Smoking-prevention programs, run by both teachers, and teachers and peers, have been introduced into school curricula in many parts of the world. This paper describes a long-term follow-up of a randomised controlled trial of a smoking education program for children conducted in Western Australia. Seven years after the first survey of 2,366 Year 7 students in 1981, 68 per cent of initial participants were traced through public records; 53 per cent of these responded to a new survey concerning smoking. Previous follow-up after one and two years had shown that both teacher-led and peer-led programs continued to reduce the taking up of smoking by girls to about the same degree, whereas in boys, the teacher-led program appeared to be effective after one year but neither program was effective after two years. In nonsmoking girls, both the intervention programs maintained their effects at the seven-year follow-up, with an almost 50 per cent reduction in smoking prevalence in the intervention group. Nonsmoking girls appeared to respond to cigarette advertisements. Mothers seemed to influence nonsmokers of both sexes and brothers seemed to influence smokers of both sexes. The seven-year follow-up confirmed the results seen at two years for boys, that the effects of the education program had dissipated. However, this study suggests that the smoking-prevention program had a lasting effect on preventing girls from taking up smoking.