This paper will deal exclusively with the relationship between diet and serum cholesterol level. Many publications have appeared in which the role of diet in the production of atherosclerosis and ischaemic heart disease has been discussed. However, in these discussions it is often not sufficiently realized that ischaemic heart disease is a condition of multifactorial origin and that diet and serum cholesterol level are only one of the several factors which influence its prevalence, incidence and effect on mortality rates.As a further restriction this paper will deal only with experience in the human. I n this respect two main methods have been used to elucidate the effect of different dietary factors on the serum cholesterol level. Different diets have been given to healthy human volunteers and in-patients, and as a result of this work it is now agreed that certain nutrients, especially saturated fats, cholesterol and sucrose (when taken in excessive amounts), increase the serum cholesterol level. There is less unanimity of opinion about the role of other dietary factors. Several investigators have shown that when saturated fat in the diet is replaced isoenergetically by polyunsaturated fat the serum cholesterol level is reduced; monounsaturated fat has a similar effect but it is less pronounced. From these observations, it has been concluded that polyunsaturated fat has a specific effect in lowering the serum cholesterol.Other authors have confirmed these observations but have found in control experiments that when saturated fat is replaced by foodstuffs like bread, rice or oats, the serum cholesterol level is also lowered. I n their opinion therefore dietary saturated fat has the specific effect of increasing the serum cholcsterol level; the replacement of saturated fat by almost all other nutrients (and not specifically by polyunsaturated fat) merely reduces the serum cholesterol level to its physiologically normal level.A second approach in assessing the role of dietary factors in determining the serum cholesterol level has been the use of what may be called nutritional epidemiology. This method compares the serum cholesterol levels of certain population groups who eat different diets, and attempts to correlate differences in serum cholesterol levels with certain factors in the diet. I n this respect special attention has been given to population groups which are distinguished by low average serum cholesterol levels, because these groups also have low frequencies of ischaemic heart disease. It is therefore of great medical importance to identify the factors which are responsible for these low serum cholesterol values. I have been able to examine three such https://www.cambridge.org/core/terms. https://doi