OBJECTIVES: To assess the impact of socio-economic status on the relationship between type of alcohol and all-cause mortality. DESIGN: A prospective population study. SETTING: The Copenhagen City Heart Study, Denmark. SUBJECTS: A total of 14,223 men and women participated in the first examination of The Copenhagen City Heart Study in 1976-1978. The participants were followed up until 18th of September 2001 during which 7208 persons died. The effect of beer, wine and spirits on mortality was stratified according to levels of education, income and cohabitation, and the association was examined after controlling for intake of the other types of alcohol, and for sex, smoking, physical activity and body mass index. MAIN OUTCOME MEASURES: Number and time of death from all causes. RESULTS: Consumers of wine were better educated and wealthier compared with beer and spirits drinkers. The association between type of beverage and mortality was noticed to differ according to socio-economic level, especially where the apparent protective effect of wine consumption tended to be strongest in the lower income and educational groups. CONCLUSIONS: This study finds the specific effects of beer, wine or spirits to moderately diverge in the socio-economic groups. Future studies addressing the association between the type of beverage and mortality may need to more thoroughly take socio-economic factors into account.