The relationship of arterial pressure (AP) to plasma arginine vasopressin (AVP) and sodium (Na) intake was determined in untreated essential hypertensive (H) and normotensive (N) subjects. The AP of H subjects averaged 147/101 mm Hg and that of N subjects, 124/79 mm Hg. Plasma AVP was elevated significantly in H subjects, averaging 8.5 pg/ml compared to 4.7 pg/ml in N subjects. Multivariant regression analysis yielded a significant correlation (r2 = 0.34) between diastolic pressure, urine Na concentration, and changes in plasma AVP. Plasma Na of H subjects averaged 2.0 mEq/liter less and urine Na concentration 22 mEq/liter less than in N subjects. Sodium intake appeared to have no influence on the plasma AVP of N subjects, but H subjects excreting Na in excess of 250 mEq/day averaged a plasma AVP twice as high as that in H subjects excreting less than 150 mEq/day. In H subjects, the influence of Na intake appeared to be related to age. In subjects less than 50 years of age, Na intake did not appear to influence chronic levels of plasma AVP, while in subjects older than 50 years who were excreting Na in excess of 250 mEq/day, plasma AVP levels were twice (13.5 pg/ml) those observed in hypertensives of the same age excreting less than 150 mEq/ day (6.5 pg/ml). The data indicate that plasma AVP tends to be elevated in moderate essential hypertension. Reduced concentrating ability of the kidneys of these subjects is suggested by decreased urine Na concentrations despite elevated plasma AVP. The observed increases of plasma AVP could be exerting a direct influence on extra- and intravascular volumes by renal and systemic vasoconstriction.