Participants in the Medical Research Council treatment trial for mild hypertension are randomly allocated to one of four treatment groups: bendrofluazide, propranolol, or a placebo for either of these drugs. The trial is single-blind. 23 582 patient-years of observation have been completed so far, 10 684 on active drugs and 12 898 on placebos. The results show an association between bendrofluazide treatment and impotence, and impotence also occurred more frequently in patients taking propranolol than in those taking placebos. Other adverse reactions significantly linked with active drugs include impaired glucose tolerance in men and women and gout in men, associated with bendrofluazide treatment, and Raynaud's phenomenon and dyspnoea in men and women taking propranolol. No corneal disease is known to have occurred in the propranolol group. Mean serum potassium level fell, and urea and uric acid levels rose, in men and women taking bendrofluazide. In the propranolol group, serum potassium and uric acid levels rose in both sexes, but the urea level rose significantly in women only.