Epidemiologic studies generally show no relationship or a weak relationship between total stroke and total blood cholesterol level. Several Japanese epidemiologic studies of hemorrhagic stroke indicate an increased risk at lower levels of blood cholesterol. However, these studies, which do not include many people with high blood cholesterol levels, do not show an increase in stroke rate at high blood cholesterol levels. The Japanese findings for hemorrhagic stroke were replicated by the Honolulu Heart Study in migrants from Japan, and in the MRFIT Screening Study. These studies also show an increase in nonhemorrhagic stroke at the highest, but not at intermediate, levels of blood cholesterol. An extensive pathologic study of the relationship of blood cholesterol to hemorrhagic and nonhemorrhagic stroke was carried out by Dr. Konishi of Osaka. The pathology was strikingly different in the two conditions, one being atherosclerotic in the larger arteries (5 mm diameter), the other arteriolosclerotic in the intracerebral arterioles (200 microns in diameter). The relationship between blood cholesterol and arterial lesions is inverse for arteriolosclerosis and positive for atherosclerosis. It is concluded that blood cholesterol influences the development of stroke above and beyond the influence of blood pressure. Furthermore, its influence is opposite for hemorrhagic and nonhemorrhagic stroke. The relationship between blood cholesterol and atherosclerotic stroke does not appear to be as strong as it is between blood cholesterol and coronary heart disease.