In an observational study designed to determine whether metabolic changes during long-term antihypertensive drug treatment are associated with an increased risk of coronary heart disease (CHD), 686 middle-aged hypertensive men recruited from a random population screening sample were followed up for 15 years. Antihypertensive treatment predominantly consisted of beta-adrenoceptor blockers and/or thiazide diuretics. CHD and diabetes mellitus were checked for at annual examinations. Time-dependent Cox regression analysis was used to determine correlation between the incidence of CHD and entry characteristics, the monitored serum levels of cholesterol and triglyceride concentrations and the development of diabetes mellitus. Univariate analysis showed the presence of diabetes mellitus a entry to the study and increased baseline serum concentrations of cholesterol and of triglycerides each to be a significant predictor of CHD, the respective relative risks (RR) being 2.12, 1.21 and 1.21. However, analysis of monitored levels of metabolic variables during follow-up showed only an increased serum cholesterol concentration to be significantly and independently associated with CHD (RR 1.07). Although serum triglyceride concentrations increased slightly during follow-up, they were unrelated to the incidence of CHD; nor was onset of diabetes mellitus during follow-up significantly associated with an increased risk of CHD (RR 1.48. Thus, the study showed the presence of metabolic disturbances such as diabetes mellitus and hyperlipidaemia before the start of antihypertensive treatment to be of positive predictive value in middle-aged hypertensive men, and an increase in the monitored serum cholesterol level to be an independent predictor of CHD, whereas neither drug-related diabetes nor an increase in the monitored serum triglyceride concentration seemed to be associated with the occurrence of CHD.