Aims To determine the effect of age, sex and cohort on the prevalence and genetic architecture of adolescent alcohol use (AAU). Design Survey study in participants registered with the Netherlands Twin Register. Setting Twins from the general population. Participants Two cohorts (data collected in 1993 and 2005-08) of twins aged 13-15, 16-17 and 18-21 years. In 1993 and 2005-08 a total of 3269 and 8207 twins, respectively, took part. Measurements Survey data on initiation and frequency of alcohol use and quantity of alcohol consumed. Findings The prevalence of alcohol initiation increased between 1993 and 2005-08 for both males and females. The largest difference was for girls observed at ages 13-15, where the prevalence increased from 59.5% to 72.4%. We also found increases in prevalence across cohorts for quantity of alcohol consumed and non-significant increases for frequency of alcohol use. From age 16 onwards, boys drank more frequently and larger quantities than girls. Genetic model fitting revealed that the genetic architecture of AAU did not differ between birth cohorts, nor were there differences between boys and girls. Genetic factors explained between 21% and 55% of individual differences in alcohol measures throughout adolescence. Shared environment explained between 17% and 64% of variance in alcohol use, across different age groups and alcohol measures. Conclusions In the Netherlands, the prevalence of alcohol initiation, frequency and quantity has increased in adolescents over a 15-year period, but there are no changes in the genetic architecture of adolescent alcohol use. © 2011 The Authors, Addiction © 2011 Society for the Study of Addiction.