During screening in the Helsinki Heart Study (HHS), a 5-year coronary primary prevention trial with gemfibrozil, some 600 dyslipidaemic individuals were detected who exhibited symptoms and signs of possible coronary heart disease (CHD). These subjects were excluded from the primary study. To secure successful conduct in the HSS, an ancillary protocol was developed for the treatment of these individuals. Three-hundred and eleven subjects were randomized to receive gemfibrozil 600 mg twice daily and 317 subjects to receive a matching placebo over 5 years in a double-blind fashion. The end-point rate, consisting of fatal and non-fatal myocardial infarction and cardiac death, did not differ significantly between the placebo and gemfibrozil groups. The same was true for total mortality. Missing key prognostic factors, e.g. true prevalence of CHD, extent of coronary artery obstructions, degree of left ventricular dysfunction, and their distribution in the groups render the results less reliable and hence the data cannot be used to refute the thesis that treatment of dyslipidaemia in manifest CHD in successful.