International data show stronger correlations of mortality from ischaemic heart disease with per capita supply of dairy products excluding fat than with dairy fat, and of estimated lactose than with dairy fat or margarine and other processed fats (positively) and vegetable oils and fats, fat of fish or wine (negatively). Butter and cheese, which have a low content of lactose, show moderate and zero correlations, respectively. Populations with low or intermediate prevalence of adult lactose absorbers have a lower supply of dairy products excluding butter (and therefore of lactose), and a lower mortality from ischaemic heart disease, than populations with a high prevalence of absorbers. Specific national and ethnic data suggest that a diet low or relatively low in lactose, in populations with low or relatively low prevalence of lactose absorbers, is more consistently associated with protection against ischaemic heart disease than are high intakes of unsaturated fatty acids, wine, alcohol or dietary fibre. In seven countries with a high consumption of dairy products (six at least with a high prevalence of lactose absorbers), trends in ischaemic heart disease mortality appear to have reflected changes in the supply of milk (and therefore of lactose), but not consistently of butter or inversely of unsaturated fatty acids. The findings reviewed in this paper call for further investigation of the subject, epidemiologically and biochemically.