A simple correlation analysis of data for 19 Western countries confirmed previously reported findings that the rate of mortality from ischemic heart disease (IHD) varies inversely with total alcohol consumption and positively with the consumption of unfermented milk proteins. However, when milk protein consumption was held constant in a multiple-partial correlation analysis, the association between total alcohol consumption and the IHD mortality rate was reduced to non-significance. In contrast, when alcohol consumption was held constant there was little effect on the high correlation between the mortality rate and the consumption of milk proteins. Multiple regression analysis confirmed the importance of the milk factor over alcoholic beverages as a determinant of variation in IHD mortality rates. It was concluded that the inverse association between alcohol consumption and IHD mortality, observed in international comparisons, is probably largely an artifact of a negative association between alcohol and milk consumption. An hypothesis is offered respecting the component of milk which might be responsible for its apparent atherogenicity.