Alcohol is part of the normal culture for a majority of the population in western countries. Few investigators or clinicians disagree with the contention that there is a positive relationship between cardiovascular disease and mortality in the upper part of the alcohol consumption curve. No such general agreement exists when infrequent users and non-users are studied. Epidemiological evidence of the relationship between cardiovascular mortality and levels of alcohol consumption is scrutinized, with emphasis on how alcohol anamnestic data are collected, the characterization of non-users, and the authors' definition of moderate consumers. The results indicate that there is hardly any evidence to advocate moderate consumption of alcohol as a health-promoting activity. On the other hand there seems to be a positive relationship between moderation in many aspects of lifestyle and some health gain.