This randomized prospective study was designated to test the hypothesis that intraoperative and postoperative administration of a potent antiinflammatory steroid (dexamethasone) would reduce patients' postoperative pain. The hypothesis is based on the concept that inflammation is associated with a lowering of the threshold for nociceptive sense organs and that the inflammatory process is responsible, at least in part, for postoperative pain. The quantity of narcotics requested by the patients during the first 72 hours of the postoperative period was used as the measure of their postoperative pain. Patients treated with steroid who underwent lumbar laminotomy and discectomy used considerably less narcotic during the postoperative period than those not treated with steroid (P value less than 0.01). Less difference was noted in patients who underwent laminectomy, and no statistically significant difference was noted for those who underwent anterior discectomy. It is concluded that the use of an antiinflammatory steroid during and after operation significantly reduces the immediate postoperative pain after lumbar discectomy and may be useful in the postoperative management of other surgical procedures.