During 1972–1973,785 symptom-free men, aged 40–49, with mild hypertension, were randomly assigned to a 5-year controlled drug treatment study, 406 in the treatment group and 379 in the control group. The basic drug was hydrochlorothiazide, in addition methyldopa or propranolol if necessary. The mean observation time was 66 months (range 60–78). At 5 years’ follow-up there was no difference in total mortality between treated and controls, nor in cardiovascular mortality. There was a good prevention of cerebrovascular disease (0 vs. 7). Regarding coronary heart disease, however, there was an opposite trend (20 vs. 13, p > 0.10). During the trial, different drug-induced changes were observed in serum lipids, glucose and uric acid. The importance of these changes are still uncertain.