In animal models of salt-dependent hypertension, hyperfiltration is associated with a faster decline in renal function and there is evidence that in hypertensive man, increased creatinine clearance is a marker of early hypertensive nephropathy. We have studied the influence of salt intake on the glomerular filtration rate (GFR) (Creatinine Clearance) in 14 patients with mild hypertension. Each patient was studied in random order and according to a crossover design, at habitual salt intake, at high salt intake (ie habitual +50/100 mmol/day) and at low salt intake (habitual -50/100 mmol/day). Protein, calcium and potassium intake was fixed across the three study periods. The control group was formed by seven healthy subjects. High salt intake, caused a significant (P < 0.01) increase in 24 h mean arterial pressure (MAP) and the expected suppression in plasma renin activity (PRA) and in plasma aldosterone. Seven patients were classified as salt-sensitive. The GFR was significantly higher (P < 0.01) at high salt intake (125 +/- 10 ml/min) than at habitual (113 +/- 7 ml/min) and at low salt intake (97 +/- 6 ml/min). On aggregate urinary salt excretion was significantly related with the GFR (P < 0.01 by correlation analysis for repeated observations) and the slope of this relationship predicted that a 100 mmol/day increase in salt intake is associated with the 14.6 ml/min rise in the GFR. The relationship between GFR and 24 h urinary salt in salt sensitive patients did not differ from that in salt resistant patients. The GFR response to salt loading was largely independent of the renin-aldosterone system. No change in arterial pressure nor in GFR was observed in healthy subjects. At fixed protein intake, changes in salt intake in the physiological range are associated with important GFR variations in mild hypertensives. As long as hyperfiltration in mild hypertension is a predictor of renal function deterioration, high salt intake, independent of the effect of arterial pressure, could be a factor that contributes to nephronic obsolescence in patients with essential hypertension.