The hypothesis that the alleviation of chronic pain with hypnosis is mediated by endorphins was tested. Six patients with chronic pain secondary to peripheral nerve irritation were taught to control the pain utilizing self-hypnosis. Each subject was tested at 5-min intervals during four 1-h sessions for the amount of reduction of pain sensation and suffering associated with hypnosis while being given, in a random double-blind crossover fashion, an IV injection of either 10 mg naloxone or a saline placebo through an indwelling catheter. The patients demonstrated significant alleviation of the pain with hypnosis, but this effect was not significantly diminished in the naloxone condition. These findings contradict the hypothesis that endorphins are involved in hypnotic analgesia.