Background Bezafibrate has effects on lipid metabolism and haemostatic function. We undertook a double-blind, placebo-controlled intervention trial, the Bezafibrate Coronary Atherosclerosis Intervention Trial (BECAIT), to establish whether bezafibrate (200 mg three times daily) could retard or prevent the progression of atherosclerotic lesions in dyslipidaemic male survivors of myocardial infarction who were younger than 45 years at the time of the event. Methods 92 patients completed an initial 3-month period of dietary intervention and were randomly assigned to treatment with bezafibrate or placebo. Dietary intervention continued throughout the trial. Coronary angiography was done at baseline and after 2 and 5 years. 81 patients (42 bezafibrate treated and 39 placebo treated) who underwent baseline angiography and at least one posttreatment angiogram were included in the efficacy analysis. The primary endpoint was change in mean minimum lumen diameter. Findings The mean minimum lumen diameter decreased from baseline to the last angiographic assessment (2 or 5 years) by 0·06 mm (95% Cl 0·15 reduction to 0·01 increase) in the bezafibrate group and by 0·17 mm (0·33 reduction to 0·09 increase) in the placebo group. The treatment effect was therefore 0·13 mm (95% Cl 0·10 to 0·15; p=0·049). Parallel treatment effects, although not statistically significant, were observed for the secondary angiographic endpoints (mean segment diameter 0·02 mm [0·01-0·04] and percentage stenosis -3·41% [-4·00 to -2·98]). The cumulative coronary event rate was significantly lower among bezafibrate-treated than among placebo-treated patients (three vs 11 patients; p=0·02). There were significant treatment effects of bezafibrate for serum concentrations of cholesterol (-9%; p<0·001), very-low-density-lipoprotein (VLDL) cholesterol (-35%; p<0·001), serum triglycerides (-31%; p<0·001), VLDL triglycerides (-37%; p<0·001), and plasma fibrinogen (-12%; p=0·001), whereas low-density (LDL) cholesterol concentrations did not change. High density lipoprotein (HDL) cholesterol increased significantly with bezafibrate (9%; p=0·02). Interpretation The results show that bezafibrate improves dyslipidaemia, lowers plasma fibrinogen, slows the progression of focal coronary atherosclerosis, and reduces coronary events in young survivors of myocardial infarction.