The effects of placebo, propranolol, and clonidine on blood pressure and blood pressure variability were examined in 14 subjects who had moderate essential hypertension. Hydrochlorothiazide was given throughout successive 4- to 5-week periods of placebo, propranolol, and clonidine administration. During each treatment period, subjects were admitted twice to the clinical research unit for 24-hour blood pressure monitoring performed during administration of placebo, propranolol and clonidine and repeated 1 to 2 weeks later during the first 24 hours after the abrupt cessation of placebo or drug administration. Blood pressure was recorded at 15-minute intervals using an automated noninvasive recorder (Arteriosonde, Roche, New Jersey). Systolic and diastolic blood pressure readings were averaged and the standard deviation taken as the measure of long-term variability (LTV). Systolic and diastolic blood pressure in sequential overlapping blocks of seven readings were averaged and the standard deviation calculated. Short-term variability (STV) was estimated as the average of the standard deviations of the running means. During placebo administration and withdrawal, systolic and diastolic blood pressures as well as LTV and STV were unchanged. Systolic and diastolic pressures did not differ during propranolol administration from those during propranolol withdrawal or placebo withdrawal. Systolic and diastolic LTV and STV did not differ during propranolol administration and withdrawal from those observed during placebo administration or withdrawal. After clonidine withdrawal, both systolic and diastolic pressures increased to values significantly greater than during clonidine administration. These values were significantly greater than those observed after placebo withdrawal, thereby indicating rebound hypertension. After clonidine withdrawal, seven of fourteen subjects developed systolic pressure rebound; diastolic pressure rebound was observed in three patients.(ABSTRACT TRUNCATED AT 250 WORDS)