Many attempts to calculate costs caused by the use of alcohol in accordance with the cost-of-illness method have been reported in the literature. However, in a decision-making perspective and with a focus on what possible interventions to undertake, cost-benefit or cost-effectiveness studies are more useful. In this study the cost-effectiveness of advice aimed at reducing 'heavy' drinking to 'moderate' drinking is calculated. Results from controlled trials, showing the short-range effects of advice, are combined with observations from long-term epidemiological studies showing the association between alcohol consumption and total mortality. This study shows that advice from primary health care staff has a potential to be a very cost-effective means of intervention. The crucial point seems to be the number of people that makes durable changes in consumption. If about 1% make lasting changes a brief intervention is relatively cost-effective (20,000 ECU/YLS), and if about 10% change resources will be saved in health care. Important effects such as increased quality of life and decreasing production losses are not taken into account.