Objectives. The aim of this study was to evaluate whether ultrafiltration is beneficial in patients with moderate congestive heart failure. Background. Ultrafiltration is beneficial in patients with severe congestive heart failure. Methods. We studied 36 patients in New York Heart Association functional classes II and III in stable clinical condition. Eighteen patients (group A) were randomly selected and underwent a single session of ultrafiltration (venovenous bypass, mean [+/- SEM] ultrafiltrate 1,880 +/- 174 ml, approximately 600 ml/h) and 18 (group B) served as control subjects. Results. Two patients in group A and three in group B did not complete the 6-month follow-up study. In group A, soon after ultrafiltration there were significant reductions in right atrial pressure (from 8 +/- 1 to 3.4 +/- 0.7 mm Hg, pulmonary wedge pressure (from 18 +/- 2.5 to 10 +/- 1.9 mm Hg) and cardiac index (from 2.8 +/- 0.2 to 2.3 +/- 0.2 liters/min). During the follow-up period, lung function improved, extravascular lung water (X-ray score) decreased and peak oxygen consumption (ml/min per kg) increased significantly from 15.5 +/- 1 (day - 1) to 17.6 +/- 0.9 (day 4), to 17.8 +/- 0.9 (day 30), to 18.9 +/- 1 (day 90) and to 19.1 +/- 1 (day 180). Oxygen consumption at anaerobic threshold (ml/min per kg) also increased significantly from 11.6 +/- 0.8 (day -1) to 13 +/- 0.7 (day 4), to 13.7 +/- 0.5 (day 30), to 15.5 +/- 0.8 (day 90) and to 15.2 +/- 0.8 (day 180). These changes were associated with increased ventilation, tidal volume and dead space/tidal volume ratio at peak exercise. The improvement in exercise performance was associated with a decrease in norepinephrine at rest, a downward shift of norepinephrine kinetics at submaximal exercise and an increase in norepinephrine during orthostatic tilt. None of these changes were recorded in group B. Conclusions. In patients with moderate congestive heart failure, ultrafiltration reduces the severity of the syndrome.