OBJECTIVE The cytoprotective effects of prostaglandin E1 on livers suffering from ischemia/reperfusion injury in the clinical setting are unproved. These effects were examined, focusing on inflammatory cytokine and nitric oxide metabolism. METHODS Twenty-four cirrhotic patients with hepatocellular carcinoma undergoing subsegmentectomy under ischemia induced only by Pringle's maneuver were divided into two groups (patients given prostaglandin E1 by injection and untreated controls) and postoperative results were compared. Peripheral blood was taken perioperatively and the plasma aminotransferase, cytokines and nitrate/nitrite levels of the two groups were compared. Two liver specimens were taken from each patient, one before ischemia and the other after hepatectomy, and the levels of inducible nitric oxide synthase and cytokine mRNAs and proteins were analyzed. RESULTS Although no apparent differences were recognized in postoperative complications or duration of postoperative hospital stay between the groups, the perioperative plasma aminotransferase level was significantly lower in the prostaglandin E1 group. Significant differences were also seen in interleukin-6 and nitrate plasma levels during the observation period and the interleukin-6 protein levels in the liver supernatants after hepatectomy in the two groups. In contrast, no significant differences were apparent between the interleukin-1 beta and tumor necrosis factor-alpha plasma levels of the two groups. The corrected fluorescence activities of interleukin-6 and inducible nitric oxide synthase mRNAs in the liver after hepatectomy correlated significantly. No interleukin-1 beta or tumor necrosis factor-alpha mRNAs or proteins were detected. CONCLUSIONS Prostaglandin E1 exerted hepatoprotective effects on livers suffering from ischemia/reperfusion injury, and interleukin-6 might play an important role in these effects.