The hypothesis that a causal relationship exists between insulin resistance and atherogenesis was first proposed over 23 years ago, and has given rise to a vast literature. Biological plausibility has been lent to the hypothesis by studies in which insulin has produced some effects in cell and tissue culture, and in vivo in arterial tissue, consistent with our understanding of the pathogenesis of atherosclerosis. Clinical studies demonstrating a complex interrelationship between insulin resistance-hyperinsulinaemia and established risk factors for CHD--hypertension, hypertriglyceridaemia, low HDL cholesterol levels and abdominal obesity--are reviewed. A review of the studies examining an independent association between hyperinsulinaemia and coronary heart disease is presented. Cross-sectional studies in both the general population and diabetes support the relationship; however, prospective studies in the general population provide limited and inconsistent support for this hypothesis and highlight the confounding effects of blood pressure, dyslipidaemia and obesity on the effects of hyperinsulinaemia. In subjects with NIDDM and impaired glucose tolerance, prospective studies have not shown a deleterious effect of insulin treatment per se, nor have they consistently shown a significantly increased risk for those with higher endogenous insulin levels. The therapeutic implications of the evidence to date are less complex and involve weight reduction by diet and exercise, the lowering of elevated blood pressure with metabolically neutral agents, the judicious use of lipid lowering drugs and, in diabetes, the use of insulin where clinically indicated.