The mechanism by which chemical hypophysectomy may relieve intractable pain in patients with cancer was evaluated in six adult rhesus monkeys. Pain was produced by electrical stimulation of the tooth pulp, and evoked potentials were recorded from the primary somatosensory cortex (PSC), the centrum medianum of the thalamus (CM), and the midbrain reticular formation (MRF) before and after injection of absolute alcohol into the pituitary gland and again following intramuscular administration of naloxone. At autopsy pituitary ablation was found to be complete in three animals and incomplete in the other three. A marked decrease in amplitude without change in latency was observed in the PSC, CM, and MRF in all animals following tooth pulp stimulation, indicating that the hypophysectomy was equally effective regardless of the degree of destruction of the pituitary gland. Naloxone reversed the hypophysectomy-induced changes in tooth pulp-evoked potentials (TPEPs) recorded from PSC in animals with complete destruction of the pituitary but had no effect in any animals on TPEPs recorded from CM and MRF and had no effect on TPEPs recorded from PSC in animals with incomplete destruction of the pituitary. These findings are compatible with the possibility that interference with sensory pathways produced by injection of alcohol may be a causative factor in the decrease in TPEPs recorded and suggest that overproduction of endorphine activated by alcohol-induced hypophysectomy may contribute to the relief of pain in patients with cancer following chemical hypophysectorny.