Experimental studies have demonstrated regression of atheromatous lesions with diet and lipid lowering drugs. In order to confirm these results clinically, reliable angiographic methods of analysis must be developed along two lines: quantitative by consensus between independent "blinded" experts, qualitative by digitalizing radiological images. Given the reproducibility of these methods, a variation of 17 to 20% in the size of the atheromatous plaques should be required to affirm a change. Five studies have been performed in patients with atherosclerosis associated with variable degrees of hyperlipidaemia and compared with a control group. NHLBI type II: 59 out of 146 patients with type II hyperlipoproteinaemia were treated with cholestyramine for 5 years with reduction of the progression of > 50% stenosis but no evidence of regression (6%). CLAS: 80 out of 160 coronary patients were treated with cholestipol and nicotinic acid for 2 years and a reduction of progression and a regression of lesions were observed in 16% of cases. Nikkila: 28 coronary patients with hyperlipidaemia were given clofibrate or nicotinic acid for a 7 year period, stabilising the evolution but with no signs of regression. FATS: 74 of 120 coronary patients with apolipoprotein B concentrations of over 1.25 g/l were given lovastatine-cholestipol or nicotinic acid-cholestipol for 2.5 years: regression of coronary lesions was observed in 32 to 39% of cases depending on the treatment administered. Olsson: reported the same results for femoral atheroma with treatments associating fenofibrate and nicotinic acid: 20% regression and reduction of progression.(ABSTRACT TRUNCATED AT 250 WORDS)