Urinary excretion of two prostacyclin metabolites was investigated in 48 subjects: 8 controls and 40 cirrhotics (9 without ascites, 22 with ascites and preserved renal function, and 9 with functional renal failure). Urinary 6-keto-prostaglandin F1 alpha (6-keto-PGF1 alpha), believed to reflect renal prostacyclin production, was significantly increased in patients without ascites and in ascitic patients with preserved renal function, but cirrhotics with renal failure showed rates similar to controls. Excretion of 2,3-dinor-6-keto-PGF1 alpha (PGI-M), the major urinary metabolite of systemic prostacyclin, was increased in all groups of patients, including those with renal failure. A single dose of sulindac, a renal-sparing prostaglandin synthesis inhibitor, reduced PGI-M but not 6-keto-PGF1 alpha in 5 cirrhotic patients. This would be consistent with the predicted renal origin of the latter and the systemic origin of the former. Ascitic patients with high urinary excretion of PGI-M (above the median value) showed significantly lower mean arterial pressure and higher plasma renin activity and aldosterone than patients with excretion below the median. Urinary 6-keto-PGF1 alpha was higher in patients with low PGI-M. Finally, creatinine clearance corrected excretion of PGI-M, as an estimation of relative plasma levels correlates both with plasma renin activity and plasma aldosterone in the 31 subjects who presented with ascites. It is suggested that enhanced synthesis of systemic prostacyclin may influence hemodynamic changes in patients with liver cirrhosis. Overproduction of systemic prostacyclin in the absence of increased renal prostacyclin synthesis appears to be characteristic of patients with functional renal failure.