Over 1,200 middle-aged men with no apparent vascular disease participated in a 5-year multifactorial primary prevention trial, in which 612 received dietetic, hygienic and--when indicated--pharmacologic treatment for the following risk factors: hyperlipidemia, hypertension, smoking, obesity and abnormal glucose tolerance. Pharmacologic therapy included hypolipidemic agents (mainly probucol and clofibrate) and antihypertensive drugs (mainly diuretics and beta blockers). At the end of the 5 years, results in these men were compared with findings in 610 high risk and 593 low risk control subjects, none of whom had received treatment. Although intervention decreased the mean risk factor status of the treated men by 33%, their 5-year coronary incidence exceeded that of the high risk control subjects (3.1% vs 1.5%). Stroke incidence, however, was markedly reduced in the treated subjects (0% vs 1.3%). Multivariate analysis showed that the coronary events occurred in patients taking beta blockers or clofibrate, while few occurred in those receiving probucol or the diuretics. The decrease in mean serum cholesterol was 15% in men receiving only probucol, and ranged from 0% to 13% in those receiving different drug combinations, including clofibrate plus probucol (11%). Probucol also markedly decreased high density lipoprotein cholesterol levels, especially when combined with clofibrate. It is possible that adverse drug effects offset the probable benefit of an improved risk profile in the treated men, thereby explaining the greater than expected occurrence of cardiac events in this group. The probucol data, however, suggest that it may not be harmful to lower the high density lipoprotein cholesterol level when there is a significant decrease in total cholesterol as well.