In patients with prior shock and liver injury, mechanical circulatory assist still has a high risk of hepatic failure. The purpose of this study was to evaluate the effect of mechanical circulation using pulsatile or nonpulsatile blood pumps on shock organs, particularly shock liver. In 14 dogs, a shock liver model was produced by clamping the descending aorta above the diaphragm. After 60 minutes of ischemia, left atrial-femoral artery bypass (LHB) was started while the clamp remained in place. A pneumatic pulsatile pump (Toyobo; Tokyo, Japan) was used in seven dogs (Gr-PP) and a centrifugal pump (Biomedicus; Minneapolis, MN) in seven (Gr-NPP). In both groups the mean arterial pressure was maintained at 80 mmHg. The mean bypass flow was 96 +/- 14 ml/kg/min for Gr-PP, and 95 +/- 35 ml/kg/min for Gr-NPP. In both groups the bile flow and arterial ketone body ratio decreased significantly after ischemia, and recovered to normal after 120 min of LHB. There was no significant difference between the two groups in this model. The other parameters (glutamic oxaloacetic transaminase [GOT], glutamic pyruvic transaminase, lactate dehydrogenase, mitochondrial GOT) were independent of the period of ischemia and reperfusion, and there was no difference between the two groups. In conclusion, these results suggest that nonpulsatile circulatory assist may not be disadvantageous for the circulatory support and recovery of liver function in the setting of shock liver.