A prospective study, including all 1083 pupils in the final year of compulsory schooling in a municipality in northern Sweden, was performed. Ninety-seven point nine per cent of the pupils were followed up after five years. They completed a comprehensive self-administered questionnaire including questions on alcohol consumption. Among men a positive correlation between long-standing unemployment (greater than 20 weeks) and alcohol consumption at the beginning and at the end of the study, as well as the change in consumption during the follow-up period was seen. This correlation was still present when earlier alcohol consumption and socioeconomic variables were controlled for. In women there was a positive correlation between unemployment and the level of alcohol consumption, but a negative correlation between unemployment and change in consumption. When controlling for motherhood and alcohol intake at the start of the study the correlation turned positive. Unemployment among women did not seem to affect recruitment to the high consumption group. Our results support the suggestion that unemployment may be a risk indicator for increasing alcohol consumption among young people, particularly in young men.