The total body production of prostacyclin was shown to be increased in cirrhotic patients, suggesting that its synthesis by blood vessels of the systemic circulation is enhanced. However, the mechanism by which the synthesis of systemic prostacyclin is stimulated is not known. The present study investigated the urinary excretion of 2,3-dinor-6-keto-PGF1 alpha, an index of total body prostacyclin synthesis, first, in cirrhotics with portal hypertension (n = 19) as compared with cirrhotics with reduced portal pressure after portacaval shunt surgery (n = 18) and with control noncirrhotic subjects (n = 11), and; second, in cirrhotics before and after intestinal decontamination by oral nonabsorbable antibiotics (n = 9 antibiotic treated patients, n = 10 control nontreated cirrhotics). Control noncirrhotic subjects showed lower urinary excretion of 2,3-dinor-6-keto-PGF1 alpha than both groups of cirrhotics (P less than 0.001). Interestingly, urinary excretion of 2,3-dinor-6-keto-PGF1 alpha was significantly higher in cirrhotics with portacaval shunt than in those with portal hypertension (P less than 0.01). The urinary excretion of 2,3-dinor-6-keto-PGF1 alpha decreased significantly after intestinal decontamination in the antibiotic-treated group (580.1 +/- 232.4 vs. 431.2 +/- 219.2 pg/mg creatinine; P less than 0.05) but not in nontreated patients (543.9 +/- 214.4 vs. 581.2 +/- 281.4 pg/mg creatinine; P = NS). These data suggest that the increased urinary excretion of 2,3-dinor-6-keto-PGF1 alpha observed in cirrhotics is not directly related to portal hypertension itself but to portal blood factors that bypass the liver. Some such factors may be of intestinal bacterial origin.