Pressure natriuresis, defined as the relationship between sodium excretion and mean arterial pressure (MAP), was assessed during graded reduction of arterial pressure with nitroprusside in 16 essential hypertensives (EH). In all patients, sodium excretion fell linearly with reductions in arterial pressure (r greater than 0.68; p less than 0.05). The per cent change of sodium excretion from control per mmHg change in MAP (delta UNaV/delta MAP) was less in patients with resting MAP above 120 mmHg than in those with lower blood pressure (1.4 +/- 0.1 versus 3.0 +/- 0.3; p less than 0.001), but the pressure at which urine flow extrapolated to zero (75 +/- 3 mmHg) was not significantly different in the two groups. The slope of the relationship between sodium excretion and arterial pressure was significantly correlated with resting MAP (r = -0.67; p less than 0.05) and with plasma volume (r = 0.61; p less than 0.05). Furthermore, the ratio delta UNaV/delta MAP also rose in concert with body fluid volumes when a salt load of 8 g of sodium chloride was added to the regular diet. Thus, the sensitivity of pressure natriuresis was determined by level of resting arterial pressure and body fluid volumes. These experiments suggest tht two mechanisms might be activated in EH to avoid dangerous sodium and volume depletion: 1) attenuation of pressure natriuresis at higher levels of arterial pressure, and 2) blunting of pressure natriuresis by volume contraction. By this hypothesis, the lower slope of pressure natriuresis is secondary to hypertension rather than its cause.