Peripheral afferent neuronal barrage from tissue injury produces central nervous system hyperexcitability which may contribute to increased postoperative pain. Blockade of afferent neuronal barrage has been reported to reduce pain following some, but not all, types of surgery. This study evaluated whether blockade of sensory input with a long-acting local anesthetic reduces postoperative pain after the anesthetic effects have dissipated. Forty-eight patients underwent oral surgery with general anesthesia in a parallel group, double-blind, placebo-controlled study. Subjects randomly received either 0.5% bupivacaine or saline intraoral injections, general anesthesia was induced with propofol, a non-opioid anesthetic, and 2-4 third molars extracted. Subjects were assessed at 24 and 48 h for postoperative pain and analgesic intake. Blood samples were collected at baseline, intraoperatively and at 1-h intervals postoperatively for measurement of beta-endorphin as an index of CNS response to nociceptor input. Plasma beta-endorphin levels increased significantly from baseline to the end of surgery in the saline group in comparison to the bupivacaine group (P < 0.05), indicating effective blockade of nociceptor input into the CNS by the local anesthetic. Pain intensity was not significantly different between groups at 24 h. Pain at 48 h was decreased in the bupivacaine group as measured by category scale and graphic rating scales for pain and unpleasantness (P < 0.05). Additionally, subjects in the bupivacaine group self-administered fewer codeine tablets for unrelieved pain over 24-48 h postoperatively (P < 0.05). These data support previous animal studies demonstrating that blockade of peripheral nociceptive barrage during and immediately after tissue injury results in decreased pain at later time points. The results suggest that blockade of nociceptive input by administration of a long-acting local anesthetic decreases the development of central hyperexcitability, resulting in less pain and analgesic intake.