Hepatic blood flow was investigated in two groups of eight anaesthetized dogs during and after one hour of either pulsatile or non-pulsatile cardiopulmonary bypass (CPB). Mean perfusion pressure was maintained at 60 mmHg. Hepatic arterial (HA) and portal venous (PV) blood flows were measured using electromagnetic flow probes, and hepatic O 2 consumption determined. The results demonstrate that: (a) pulsatile CPB reduces peripheral vascular resistance during and after perfusion, and more effectively preserves pump flow rate and cardiac output than non-pulsatile CPB; (b) total liver blood flow is sustained more effectively by pulsatile CPB than by non-pulsatile CPB due to relative preservation of both HA and PV flows; (c) hepatic O2 consumption is only marginally better preserved during and after pulsatile CPB than with non-pulsatile perfusion. We conclude that: (a) pulsatile CPB tends to maintain hepatic blood flow through a relative reduction in HA vascular resistance and an improvement in PV flow produced passively by a greater pump flow rate; (b) pulsatile CPB less effectively benefits hepatic O2 consumption because of poor O2 uptake from the hepatic PV blood supply.