Synthetic alpha-human atrial natriuretic peptide (alpha-hANP), 1 micrograms/kg, was intravenously given to 16 cirrhotic patients with ascites and 9 control subjects (CS) to investigate major factors responsible for sodium retention and refractory ascites. The following parameters were measured before and after alpha-hANP administration; such as lithium clearance (CLi) as an index of fluid delivery to the distal tuble, mean arterial pressure (MAP), urinary sodium excretion rate (UNaV), urine volume (V), glomerular filtration rate (GFR), effective renal plasma flow (ERPF), plasma renin activity (PRA), plasma aldosterone concentration (PAC), urinary excretion of prostaglandin (PG)E2, 6-keto-PGF1 alpha (6-k-PGF1 alpha), and thromboxane B2 (TxB2). Patients were divided following alpha-hANP administration into 2 groups as "good responders (GR)" and "poor responders (PR)", in which GR was defined as the group showing 2-fold-increase in UNaV. In contrast, PR had significant lower MAP (71.8 +/- 5.04 mmHg), GFR (21.3 +/- 3.90 ml/min), ERPF (158.0 +/- 43.8 ml/min), FELi (CLi/GFR; 12.6 +/- 1.26%), and higher PRA (8.72 +/- 0.99 ng/ml/h) and PAC (12.2 +/- 3.13 ng/dl) than GR. GR demonstrated almost same natriuretic response as CS with an increase of GFR and renal PGs synthesis, and a decrease of FELi despite reduction in blood pressure. However, alpha-hANP did not suppress PRA, PAC, and distal tubular reabsorption of sodium (FDRNa = 1-FENa/FELi) in cirrhotic patients, whereas suppressed in CS. UNaV correlated with FELi (r = 0.687, p = 0.01) and GFR (r = 0.777, p = 0.01). PRA correlated with FELi r = 0.669, p = 0.015), GFR (r = -0.634, p = 0.018), and MAP (r = 0.858, p = 0.001) only in cirrhosis. These results therefore indicated that hypotension caused by hemodynamic alteration and extremely stimulated renin release might effect on proximal tubular sodium reabsorption and GFR, leading to sodium retention and diuretic resistance in cirrhosis.