We compare the efficacy of a tubular pulsatile pump and a conventional volumetric pump (IVAC 571), connected to a neonatal hemofiltration circuit with an FH22 filter, for continuous renal replacement therapy in 54 Maryland pigs weighing 8-16 kg. Three different flow rates (30 ml/min in 12 cases, 15 ml/min in 22 cases, and 5 ml/min in 20 cases) were used over a 2-hour period. Hemofiltration and hemodiafiltration were performed, and measurements of ultrafiltrate flow, circuit pressures, heart rate, blood pressure, temperature, urea, creatinine, total proteins, Na, K, Cl, hematocrit, and hemolysis parameters (aspartate transaminase, lactic dehydrogenase, haptoglobin, indirect bilirubin, free hemoglobin) were made. There were no differences in ultrafiltrate flow between the two pumps. Ultrafiltrate volume was significantly higher with higher flows (p < 0.01). The technique was well tolerated by all pigs. When each blood flow was analyzed separately, cross-filter pressure drop was significantly higher in the volumetric pump than in the tubular pulsatile pump (p < 0.05). No significant differences in heart rate, blood pressure, or analytical determinations were seen between the two pumps. We conclude that pulsatile and volumetric pumps can be uses as an alternative to roller pumps for continuous venovenous renal replacement therapy in neonates and infants.