The Paris Prospective Study is a long-term investigation of the incidence of coronary heart disease in a large population of working men. The first follow-up examination involved 7,038 men, aged 43-54 years. Subjects with impaired glucose tolerance or diabetes (n = 943) were selected from the total population for a separate analysis of coronary heart disease mortality risk factors. During a mean follow-up of 11 years, 26 of these 943 subjects with abnormal glucose tolerance died from coronary heart disease. Univariate analysis showed that plasma triglyceride level (p less than 0.006), plasma cholesterol level (p less than 0.02), and plasma insulin level both fasting and 2-h post-glucose load (p less than 0.02), were significantly higher in subjects who died from coronary heart disease compared to those who did not. In multivariate regression analysis using the Cox model, plasma triglyceride level was the only factor positively and significantly associated with coronary death. The distribution of plasma triglyceride levels was clearly higher for the subjects who died from coronary heart disease compared to those who did not die from this cause or were alive at the end of the follow-up. This new epidemiological evidence that hypertriglyceridaemia is an important predictor of coronary heart disease mortality in subjects with impaired glucose tolerance or diabetes suggests a possible role of dyslipidaemia in the excessive occurrence of atherosclerotic vascular disease in this category of subjects.