The morbidity and mortality benefits of new forms of continuous renal replacement therapy remain controversial. The authors have compared a cohort of consecutive prospectively studied critically ill patients with acute renal failure treated with continuous venovenous hemodiafiltration (CVVHD) (n = 76) to a previously described antecedent group of patients treated in intensive care with intermittent hemodialysis or peritoneal dialysis (conventional dialysis [CD]) (n = 84). Patients were comparable for mean age, gender distribution, and mean number of failing organs (CVVHD: 4; CD: 3.9). CVVHD patients were more severely ill as measured by APACHE II score (CVVHD: score of 29; CD: score of 25.8). Despite their greater illness severity, CVVHD patients more often survived to hospital discharge (CVVHD: 40.8%) than did CD patients (CD: 29.8%; NS). After adjustment for illness severity, in patients with two, three, or four failing organs, survival was 54.3% for CVVHD versus 29.3% for CD (p < 0.01). Survival was 48% for CVVHD patients with an intermediate APACHE II score (24 to 29), compared with 12.5% for comparable CD patients (p < 0.01). No statistically significant differences were seen at either extreme of illness severity. Complications were significantly fewer during CVVHD (1 vs. 18). These data support the view that CVVHD reduces morbidity and mortality in critically ill patients with acute renal failure.