One of the main objectives of antihypertensive therapy is to preserve renal function from the deleterious effects of elevated blood pressure. Diuretics alone or in combination are effective for the treatment of arterial hypertension. Nevertheless, their use is accompanied by unwanted biochemical side effects, which have been attributed to their renal effects. During the last 10 years a group of 211 patients, diagnosed as having essential hypertension, were followed up. During the follow-up, they received a stepped-care therapeutic regimen consisting of nonpharmacologic measures (group 1), hydrochlorothiazide and amiloride (group 2), propranolol (group 3) and, if necessary, hydralazine (group 4). During the study, blood pressure remained within comparable, well-controlled levels in the 4 groups of patients. A progressive elevation of the levels of total serum cholesterol and glucose was observed in every group. The elevation attained statistical significance (p less than 0.01) after 4 years of therapy in those groups receiving the diuretic alone or in combination. Nevertheless, after 8 years of follow-up, the increment observed in these 2 parameters did not differ when patients in group 1 were compared with those in the remaining groups, indicating that thiazide diuretics could contribute to the earlier appearance of forthcoming events. Serum potassium levels were significantly lower (p less than 0.01) in groups 2 and 3 than in group 1. At the same time, we have observed the progressive appearance of clinically relevant proteinuria in 15.2% of patients, and the range of protein excretion ranged from 350 to 3,700 mg/24 hours. The appearance of proteinuria did not depend on the lack of control of blood pressure, nor on the different therapeutic requirements but was accompanied by a progressive decrease in creatinine clearance. The consequences of the renal effects of diuretics are of great importance during long-term therapy. The present results indicate that diuretics preempt the appearance of a forthcoming increase in serum glucose and cholesterol, and lessen the clinical relevance of these events.(ABSTRACT TRUNCATED AT 250 WORDS)