Abstract. Clinical and experimental evidence suggests that glucocorticoids may be effective in the treatment of neuropathic pain, but their mechanism of action is unknown. We gave triamcinolone (3 mg / kg) to rats with an experimental post-traumatic painful peripheral neuropathy, chronic constriction injury (CCI), five days after nerve injury, when the abnormal pain syndrome is known to be present; and pain sensitivity was measured on postoperative days 7 -14, a period during which symptoms are known to be at approximately peak severity. Additional CCI rats were treated similarly; and then they were sacrificed five days after the injection for an immunocytochemical analysis of endoneurial tumor necrosis factor-alpha (TNFα), macrophages, and mast cells in the sciatic nerve proximal to the site of injury. Vehicle-injected CCI rats demonstrated the expected neuropathic pain symptoms. Triamcinolone-treated CCI rats had a statistically significant reduction in the magnitude of heat-hyperalgesia and mechano-allodynia, but there was no effect on cold-allodynia or mechano-hyperalgesia. On the nerve-injured side of vehicle-injected rats, TNFα was present in Schwann cells and mast cells. On the nerve-injured side of triamcinolone-treated rats, there was a significant (71.5%) reduction in the number of TNFα-positive mast cells. Our results suggest that glucocorticoid therapy for neuropathic pain may work via the reduced expression of TNFα in endoneurial mast cells.