OBJECTIVE To study the tubular site or sites of the natriuretic action of atrial natriuretic factor and the possible differences between healthy subjects and patients with essential hypertension. METHODS Nine healthy volunteers and six patients with essential hypertension were studied on four test days under standard conditions, water loading and hydropenia with mannitol or saline loading. On each test day baseline, atrial natriuretic factor infusion (1 microgram/min) and recovery measurements were performed after equilibrium had been reached. The measurements included the atrial natriuretic factor (ANF) plasma levels, blood pressure, glomerular filtration rate (GFR), effective renal plasma flow, urinary osmolality and the (fractional) excretions of (free) water, sodium and potassium. METHODS Fractional free-water excretion and free-water reabsorption (as a function of osmolar clearance) were calculated during water loading and hydropenia with mannitol or saline loading, respectively, using standard formulae. [125I]-iothalamate and [131I]-hippuran were infused continuously for the measurement of GFR and effective renal plasma flow, respectively. RESULTS The plasma ANF concentration rose five- to eightfold during the infusion of ANF, which induced an increase in urinary sodium excretion, a small increase in GFR and a decrease in the effective renal plasma flow. Moreover, ANF induced an increase in fractional free-water excretion and a decrease in fractional free-water reabsorption. These changes did not correlate with changes in GFR. The blood pressure and potassium excretion were not affected. The effects of ANF on the plasma ANF levels, natriuresis and renal haemodynamics did not differ between the normotensive and the essential hypertensives. However, the increase in fractional free-water excretion was significantly greater in the patients with essential hypertension and correlated significantly with blood pressure (r = 0.56, P < 0.05). CONCLUSIONS These results indicate that the infusion of ANF at a low dose induces a similar natriuretic response in normotensive subjects and in patients with essential hypertension. This natriuresis is probably the result of both a glomerular and a tubular effect of ANF. Proximal as well as distal tubular sites seem to be involved. In essential hypertension an enhanced proximal as well as an impaired distal tubular action of ANF can be hypothesized.