This study was designed to determine whether blood pressure reduction, per se, causes adverse effects on cognitive and behavioral function in elderly hypertensive patients. Men with mild-to-moderate diastolic hypertension who had passed their 60th birthday were entered into the trial. After a placebo washout period, they were assigned in a randomized, double-blind manner to one of two groups receiving hydrochlorothiazide (either 25 mg once or twice daily or 50 mg once or twice daily). Responders entered a 1-year maintenance period. Nonresponders were randomly assigned to double-blind treatment with hydralazine, methyldopa, metoprolol, or reserpine added to the diuretic therapy. During the placebo and treatment periods, patients underwent a battery of psychometric tests designed to assess cognitive function, motor skills, memory, and affect. A separate questionnaire assessed the patient's ability to perform activities of daily living. A subset of patients blindly being treated with placebo received the same battery of tests as a control for practice effect. The results showed that there was similar improvement on the psychometric tests between those patients whose blood pressure was successfully reduced and the placebo-treated control group. Therefore, the practice effect did not obscure a true deterioration in function. There were no substantive differences between the lower and higher doses of diuretic or among the four drugs added to the diuretic, although there were qualitative differences in side effects. We conclude that blood pressure reduction, per se, does not adversely affect cognitive and behavioral function in elderly hypertensive patients and that antihypertensive treatment is safe and effective in these patients.