Mechanisms underlying the pain state in humans that follows incomplete injury to peripheral nerve are little understood. To gain better understanding of this phenomenon, this study evaluated the effects on the thermally evoked hind-paw withdrawal latency produced by the intrathecal administration of morphine, U-50 488H (U-50), (D-Pen2, D-Pen5)-enkephalin (DPDPE), ST-91, baclofen, muscimol, and 5'-N-ethylcarboxamide-adenosine (NECA) in normal rats and in rats with a hind paw rendered unilaterally hyperesthetic by the unilateral application of loose ligatures to the sciatic nerve. In the animals with one ligated nerve, the hind-paw latency for the ligated paw was typically 2-4 s less than that for the nonligated paw, at 7-11 days postoperatively. In normal rats prepared with chronic intrathecal catheters, dose-dependent increases in paw withdrawal latency were observed; the order of activity was: baclofen, ST-91, morphine, muscimol, DPDPE much greater than U50, NECA greater than or equal to 0. In the nonligated (nonhyperesthetic) paw of the lesioned animals, intrathecal agents also resulted in a dose-dependent increase in the paw withdrawal latency; the order of potency was: NECA, baclofen, morphine, ST-91, muscimol, DPDPE greater than U50 greater than or equal to 0. For both NECA and morphine, the median effective dose (ED50) values were significantly less in the nonhyperesthetic hind paw. For the hyperesthetic paw, the dose-response curves were parallel to those obtained concurrently in the nonhyperesthetic paw but were shifted significantly to the right by a factor of 3-5, with the rank order of activity in the hyperesthetic paw being baclofen, morphine, muscimol, DPDPE greater than ST-91, NECA, U50 greater than or equal to 0. These data indicate that 1) spinal receptor systems that alter thermal afferent processing in the normal animal are similarly active in the hyperesthetic paw of the lesioned animal; and 2) unexpectedly, despite similar predrug response latencies, certain receptor systems regulating the response in the nonhyperesthetic paw of the lesioned rat (morphine and NECA) show greater activity than in the nonlesioned rat.