OBJECTIVE: To investigate uptake of smoking in a cohort of 11 to 12 year olds related to awareness of advertised cigarette brands named. DESIGN: Self completed questionnaires administered to whole classes of schoolchildren in June 1993 and June 1994. SETTING: Primary, middle, and secondary schools in the north and south of England. SUBJECTS: 1450 pupils aged 11 and 12 years at the time of the first survey. MAIN OUTCOME MEASURES: Onset of smoking and brands smoked by the second survey related to cigarette brands named in the first one. Less advertised brands were used as the base for calculating odds ratios. RESULTS: Girls who named the most advertised brands-namely, Benson and Hedges alone (odds ratio = 2.50, 95% confidence interval = 1.18 to 5.30) or Benson and Hedges and Silk Cut (2.15, 1.04 to 4.42) in the first survey were at greatest risk of taking up smoking by the second one. The difference was similar but not significant for boys. Boys and girls who named the least advertised brands in the first survey were at no greater risk of taking up smoking by the second survey than those who named no brands (boys odds ratio = 0.49 (0.24 to 1.01); girls 0.79 (0.38 to 1.62)). New smokers were more likely to smoke any available brand (29.5%) or a less advertised brand such as Embassy (24.6%) than the most advertised ones, Benson and Hedges (19.7%) and Silk Cut (14.8%). Established smokers were more selective, only 15% smoking any available brand and 38.3% smoking Benson and Hedges. CONCLUSIONS: Cigarette advertising appears to increase children's awareness of smoking at a generic level and encourages them to take up the behaviour, beginning with any cigarettes which are available and affordable.