Cancer pain treatment is well established. The World Health Organization provides clinicians an "analgesic ladder" scheme to optimize cancer pain treatment. At the beginning of the pain treatment, oral analgesic administration is preferred. The analgesic dose must be individualized. Many published papers describe the spinal administration of opioids in combination with various other drugs such as bupivacaine in selected patients with cancer pain. Although complications have been reported to be few, some recent reports debate this idea. We first describe a population of 92 cancer patients, 13 of whom received intrathecal morphine. We then present our experience with a separate group of 33 cancer patients who were also managed using intrathecal morphine. Based on this experience, the generally accepted indications for the technique appeared to be justified. Concern about spinal infection is well considered, however. Three out of those patients developed meningitis, a complication rate that is far too high.