How effective are peer-led programmes in preventing the uptake of smoking by children? In 1981, we conducted a randomized controlled trial of a school-based educational programme for the prevention of smoking in children who were in their seventh year at school. In this article, the reported results of two years of follow-up confirm an earlier report that both teacher-led and peer-led programmes resulted in a reduction, to about the same degree, in the uptake of smoking by girls, while only the teacher-led programme appeared to be effective in boys. In girls, both the teacher-led and peer-led programmes maintained their effects over the two years of follow-up with adjusted differences in prevalence rates of the uptake of smoking relative to the control group of -6.6% (95% CL, -17.3%, 4.0%) and -8.1% (95% CL, -18.9%, 2.7%), respectively, after two years. In boys, the effect of the teacher-led programme was reduced substantially by the second year with a difference in the prevalence rate of -2.8% (95% CL, -11.2%, 5.6%); for the peer-led programme the difference in the prevalence rate was +6.4% (95% CL, -3.6%, 16.4%). Other variables which had a significant effect on the smoking behaviour were the perceived response to cigarette advertising, parental and sibling smoking status, the perceived parental sanctions on smoking behaviour, selected peer influences and the intention to smoke. The children's perceived responses to cigarette advertising showed the strongest and most consistent evidence of an effect on the uptake of smoking by children who initially were non-smokers. After adjustment for the effects of other variables there was an excess of 15.0% (95% CL, 2.1%, 27.9%) in the prevalence rate of smoking after two years for girls who thought that they were influenced by advertising compared with those who did not. The corresponding difference for boys was 15.3% (95% CL, 4.0%, 26.6%). As smoking-prevention programmes only may delay the onset of smoking in children, it is important that legislative measures be introduced to reduce the effects of cigarette advertising.