Genetic and environmental determinants of self-reported alcohol consumption were investigated in a sample of 2513 twin pairs who were first assessed at age 16 and were followed-up at age 17. At age 16, 77% of the sample was drinking, and 65% of drinkers reported drinking to intoxication. Both drinking and drinking to intoxication increased at the 1-year follow-up. Model fitting indicated that most of the variance in drinking initiation was due to shared environmental effects but that shared environmental effects were less important, and additive genetic effects were more important, in explaining frequency of drinking among subjects who had already initiated drinking. Similarly, shared environmental effects explained most of the variation in initiation of drinking to intoxication but were less important in explaining frequency of intoxication among subjects who had already initiated drinking to intoxication. The magnitude of genetic and environmental estimates for males and females did not differ significantly, but it was clear that either different genetic factors or different shared environmental factors were influencing males and females. For all drinking variables studied, shared environmental effects decreased from age 16 to age 17, while additive genetic effects increased from age 16 to age 17.