The relation between nutrient intake and progression of coronary artery disease was examined in 50 men receiving a lipid-lowering diet or usual care in the St. Thomas' Atherosclerosis Regression Study. Nutrient intake was assessed by diet history. Changes in coronary angiograms were measured by quantitative image analysis. In univariate linear regression analysis progression of disease over 39 months, as measured by decrease in minimum absolute width of coronary segments, was directly related to dietary energy (p < 0.001) and to the absolute intakes of total fat (p < 0.001), saturated fat (p < 0.001), monounsaturated fat (p = 0.016) and cholesterol (p = 0.06). No significant associations were seen with polyunsaturated fat, carbohydrate, protein, fiber, alcohol or with the ratio of intakes of polyunsaturated fatty acids to saturated fatty acids. In multiple linear regression analysis the associations of change in minimum absolute width of coronary segments with total or saturated fat persisted when adjusted for plasma low-density lipoprotein cholesterol concentration, age, weight, blood pressure, smoking or treatment group assignment. The findings suggest that in middle-aged men, progression of CAD is strongly influenced by intake of saturated fatty acids, an effect mediated in part by mechanisms other than the influence of this nutrient on plasma cholesterol and low-density lipoprotein cholesterol.