The aim of this study was to determine the plasma levels, cardiac release and splanchnic extraction of atrial natriuretic factor in cirrhosis with ascites. The plasma concentration of immunoreactive atrial natriuretic factor in samples obtained from an antecubital vein was measured in 18 healthy volunteers and in 35 cirrhotics with ascites. In 11 of these cirrhotics and in 11 patients admitted to the hospital for the study of a thoracic pain who had no clinical or hemodynamic signs of cardiac failure (control group), the plasma levels of immunoreactive atrial natriuretic factor in samples from the coronary sinus, right atrium, pulmonary artery, hepatic vein and femoral vein were determined and the coronary sinus blood flow measured by thermodilution. Cirrhotic patients showed significantly higher plasma levels of immunoreactive atrial natriuretic factor in each vascular territory studied than did control subjects (coronary sinus: 101.2 ± 10.6 vs. 26.1 ± 4.7 fmoles per ml; right atrium: 32.5 ± 5.8 vs. 9.4 ± 3.5; pulmonary artery: 36.8 ± 10.1 vs. 7.5 ± 2.4; hepatic vein: 10.7 ± 2.0 vs. 2.7 ± 0.8; femoral vein: 18.2 ± 2.4 vs. 3.1 ± 0.9; antecubital vein: 14.7 ± 1.6 vs. 4.0 ± 0.8). The coronary sinus blood flow was also higher in cirrhotics (200 + 22 ml per min) than in controls (105 ± 7 ml per min). Consequently, the estimated cardiac release and cardiac production of immunoreactive atrial natriuretic factor were strikingly increased in cirrhotics (13,334 ± 2,007 and 5,484 ± 1,734 fmoles per min, respectively) as compared to control subjects (1,669 ± 338 and 1,431 ± 350 fmoles per min, respectively; p < 0.01). The splanchnic extraction of immunoreactive atrial natriuretic factor was similar in both groups. These results indicate that the plasma levels of immunoreactive atrial natriuretic factor are increased in patients with cirrhosis and ascites, and that this is due to an increased release and not to an impaired splanchnic degradation of the peptide.