The coincidence of increased red blood cell mass with obesity and hypertension has been confirmed in a number of studies. Since erythropoietin (EPO) induces hypertonic effect, we aimed to study the possible involvement of endogenous EPO in the pathogenesis of obesity-related hypertension (ORH). Seventy-two non-obese (body mass index-BMI < 25kg/m2) and 47 obese (BMI > 30 kg/m2) patients with essential hypertension, and 32 non-obese normotensive control subjects were studied on 120 mmol Na/24 h diets for 3 days and 30 mmol Na/24 h diets for the following 3 days. Serum EPO levels, plasma renin activity (PRA), aldosterone (PAC) and urinary sodium and potassium excretion were estimated under both conditions. Hematocrit, hemoglobin and serum iron, ferritin and TIBC were measured in basal conditions. Patients with ORH showed significantly higher hematocrit values, hemoglobin levels and erythrocyte counts (p < 0.05) than non-obese essential hypertensives and controls, although no differences in serum EPO levels (13.6 ± 1.1 and 12.8 ± 1.0 mU/ml, 15.3 ± 1.2 and 13.3 ± 1.1, 12.2 ± 2.0 and 10.95 ± 1.7, respectively) and log EPO X %Hct product were found between groups. In both hypertensive groups, BMI but not blood pressure positively correlated with hematocrit. A negative correlation between changes in EPO and PRA induced by sodium restriction was found in lean hypertensive subjects only (r = -0 .42, p < 0.01). This finding suggests that in contrast to non-obese patients with essential hypertension, in subjects with ORH activity of the renin-angiotensin-aldosterone system does not seem to be related to EPO secretion. Since in both lean and obese essential hypertensive subjects EPO and log EPO X %Hct product are within normal range, the contribution of endogenous EPO to the pathogenesis of hypertension in these patients is unlikely. (H ppertens Res 1994; 17: 43-48)