Inoculation of the right hind paw with Mycobacterium butyricum rapidly led to swelling and inflammation. The afflicted limb showed an enhanced sensitivity to noxious pressure (hyperalgesia) and a reduced sensitivity to noxious heat 24 h following treatment. Both naloxone and MR 2266 (which has greater activity at kappa-opioid receptors) further increased the sensitivity to pressure (that is, potentiated the hyperalgesia) but did not affect the response to heat. They did not affect the response of the uninflamed paw. At 1 week, only MR 2266 was effective. At both 24 h and 1 week, the inflamed paw showed pronounced supersensitivity to the antinociceptive action of morphine against noxious pressure. At both 24 h and (to a greater extent) 1 week, a rise in levels of immunoreactive (ir)-dynorphin (DYN) was seen in the ipsilateral dorsal horn of the lumbar spinal cord. There was no alteration in the contralateral dorsal horn or in either ventral horn. Furthermore, levels of ir-met-enkephalin (ME) and ir-leu-enkephalin (LE) were unaffected. There was no difference in the density of mu-, delta- or kappa-binding sites in any part of the lumbar cord, at either 24 h or 1 week, between ipsilateral and contralateral tissue. By 3 and 5 weeks postinoculation, the symptoms had spread to the contralateral hind limb and ir-DYN was elevated in the contralateral dorsal horn and the ipsilateral ventral horn. At 5 weeks, levels of ir-ME and ir-LE also were increased in the ipsilateral and contralateral dorsal horns, but not in the contralateral ventral horn. Furthermore, levels of ir-DYN were increased in the cervico-thoracic spinal cord, and rats displayed adrenal hypertrophy and a rise in plasma levels of ir-beta-endorphin (beta-EP). These data indicate: (1) Peripheral inflammation localized to a single limb selectively modifies levels of ir-DYN in ipsilateral dorsal horn. The effect is specific to DYN as compared to ME and LE. The density of mu-, delta-, or kappa-receptors in the lumbar spinal cord is unmodified. (2) The altered response to opioid agonists and antagonists shown by rats with an inflamed limb may be selective to the injured tissue. (3) Alterations in opioid systems associated with unilateral hind limb inflammation may not be exclusively chronic in nature: they appear very rapidly (within 24 h) of the induction of pain. With time, the contralateral limb becomes affected and, eventually, the effects resemble those seen with generalized polyarthritis.