Aims. To evaluate the effects of changes in aggregate alcohol consumption on overall accident mortality in 14 western European countries after 1950, and to compare traditional beer, wine, and spirits countries with respect to the impact of alcohol. Design, setting and participants. The countries were sorted into three groups - traditional spirits countries of northern Europe, traditional beer countries of central Europe and wine countries of southern Europe. Gender- and age-specific annual mortality rates were analysed in relation to per capita alcohol consumption, utilizing the Box-Jenkins technique for time series analysis. All series were differenced to remove long-term trends. The results of the analyses in individual countries were pooled within each group of countries to increase the statistical power. Measurements. Overall accident mortality data for 5-year age groups were converted to gender and age specific mortality rates in the age groups 15-29, 30-49 and 50-69 years. Rates were age adjusted within groups. Data on per capita alcohol consumption were converted to consumption per inhabitant 15 years and older. Findings. The analyses demonstrated a statistically significant and positive relationship between changes in aggregate alcohol consumption in all three groups of countries. The estimated effect parameter was larger in northern Europe than in central Europe, and smallest in southern Europe. Conclusion. The results are compatible with the hypothesis that accident mortality rates are influenced by per capita alcohol consumption in southern, central and northern Europe. However, alcohol appears to play a larger role in northern Europe than in southern Europe.