Epidemiologic techniques allow us to use “natural experiments” to investigate the conditions that lead to disease. This article reviews the critical early observations about the epidemiology of coronary heart disease (CHD) and illustrates the key observations with data from five countries. These data suggest that only populations that consume a diet that is high in saturated fat or milk products develop high rates of CHD. Both smoking and hypertension promote the disease in the presence of an atherogenic diet but are insufficient by themselves to cause epidemic CHD. Both France and Japan have disease rates that are lower than would be predicted from the cholesterol, blood pressure, and smoking levels in their populations. The antioxidant effects of vegetables and the paucity of dairy products appear to be the factors that are protecting the French population. The platelet-inhibiting effects of marine oils appear to be providing protection to the Japanese population.