Lipid-lowering therapy ameliorates coronary atherosclerosis in patients with raised concentrations of low-density-lipoprotein (LDL) cholesterol. We have investigated whether a similar benefit can be obtained in normocholesterolaemic patients. We studied 79 normocholesterolaemic patients with coronary heart disease (70 male, 9 female), mean age 58 years, all non-smokers, with mean total cholesterol concentration 5.5 mmol/L. All patients received diet therapy and were randomly assigned placebo (39) or active treatment (40) with pravastatin, nicotinic acid, cholestyramine, and gemfibrozil stepwise as needed to reach the specified goal (total cholesterol < or = 4.1 mmol/L, ratio of LDL/high-density-lipoprotein [HDL] cholesterol < or = 2.0). Coronary angiograms at baseline and after 2.5 years of treatment were analysed by computer-assisted quantitative techniques. There was no significant difference in coronary atherosclerosis during follow-up between the active treatment and placebo groups; the mean minimum diameter narrowed significantly but to the same extent in both groups (change baseline to 2.5 years 0.14 [SD 0.42] and 0.15 [0.42] mm, respectively, both p < 0.001). Similarly, the change in percentage stenosis did not differ between the groups (2.1 [10.6] vs 2.4 [10.3]%). By multiple regression analysis, the adjusted difference between the groups was a 0.04 mm (95% CI -0.04 to 0.12 mm) increase in minimum diameter and a 0% (-1.7 to 1.7) change in percentage stenosis. The groups differed significantly in plasma lipids (% change in active minus % change in placebo group: -28% total cholesterol, -41% LDL-cholesterol, 13% HDL-cholesterol, -26% triglycerides, -31% apolipoprotein B, all p < 0.001). Thus, intensive pharmacological treatment of normocholesterolaemic patients has significant effects on plasma lipid concentrations but no angiographically measurable benefit on the coronary arteries.