We examined the analgesic and anti-allodynic effects of morphine and U-50,488H (trans-(+/-)-3,4-dichloro-N-methyl-N-(2-[1-pyrrolidinyl]-cyclohexyl)-benzeneacetamide methanesulfonate salt), a selective kappa-opioid receptor agonist, and the development of tolerance to their effects in neuropathic pain model mice induced by sciatic nerve ligation (SNL). In the tail-pinch method, morphine at 10 mg/kg, s.c. produced a weak analgesic effect in SNL mice; however, U-50,488H at 5 mg/kg, s.c. produced an analgesic effect equipotent to that in normal mice. In contrast, morphine produced an adequate analgesic effect when given either intracerebroventricularly (i.c.v.) or intrathecally (i.t.), but U-50,488H only produced analgesia when given i.t. Repeated administration of morphine (either i.c.v. or i.t.) or U-50,488H (either s.c. or i.t.), did not induce tolerance to the effect. In the static allodynia test with an application of von Frey filaments, both compounds given s.c. suppressed the allodynic effect, but in the dynamic allodynia test involving lightly stroking the plantar surface with a cotton bud, only U-50,488H produced an anti-allodynic effect. Repeated administrations of both compounds did not develop tolerance to these anti-allodynic effects. Thus, U-50,488H was found to be a highly effective at blocking hyperalgesia and allodynia in nerve injury, and these findings suggest that kappa-opioid receptor agonists are attractive pharmacological targets for the control of patients with neuropathic pain.