Background. Despite improvements in surgical techniques and perioperative care, severe complications lead to long hospital stays for some esophageal cancer patients. The purpose of this study was to evaluate the safety and effectiveness of perioperative steroid therapy on the postoperative clinical course. Methods. Fifty-seven patients operated for esophageal cancer in 1997 and 1998 were treated with perioperative steroid therapy. Fifty consecutive patients operated in 1995 and 1996 served as a control group. In the steroid group, each patient was given 250 mg of methylprednisolone intravenously before operation followed by 125 mg on postoperative days 1 and 2. Serum interleukin-6, polymorphonuclear cell elastase, and C-reactive protein levels, and the postoperative clinical course were compared between the groups. Results. Morbidity rates including hyperbilirubinemia, anastomotic leakage, and liver dysfunction were significantly lower in the steroid group than in the control group. Days until extubation and hospital stay were significantly shorter for the steroid group. Inflammatory mediators, body temperature, heart rate, and respiratory index after the surgical procedure were significantly lower in the steroid group. Adverse effects possibly caused by steroid therapy were not observed. Conclusions. Perioperative steroid therapy was safe and effective for the inhibition of inflammatory mediators and the improvement of the postoperative clinical course of patients with esophageal cancer. (Surgery 2000;128:791-8.)