Sensory abnormalities and changes in spontaneous behavior were examined after a photochemically induced ischemic lesion of the rat sciatic nerve. Male adult rats were anesthetized and the sciatic nerve was exposed. After the intravenous injection of a photosensitizing dye, erythrosin B, the exposed nerve was irradiated just proximal to the nerve trifurcation with light from an argon laser. Three different irradiation times were used, 30 s, 1 and 2 min. In sham-operated rats, the exposed sciatic nerve was irradiated for 2 min without prior injection of the erythrosin B. Rats were tested for the presence of mechanical, cold and heat allodynia or hyperalgesia. All the animals in the 1- and 2-min irradiation groups developed mechanical, cold and heat allodynia after nerve irradiation. A significant dose-dependent effect of laser exposure time was observed for all modalities tested (2 min>1 min>30 s=sham). The maximum effects were observed at 3 and 7 days postirradiation and remained present for up to 10 weeks. No significant contralateral effects were observed in any of the groups. In three separate groups of rats (1, 2 and 4 min of laser exposure), the presence of possible signs of spontaneous pain (paw shaking, paw elevation and freezing behavior) was tested. A significant and exposure time-dependent increase in spontaneous paw elevation and paw shaking was observed which was maximal at week 1, but resolved at 4 weeks (4 min>2 min>1 min>sham). In addition, animals in all ischemic groups, but not in the sham group, showed a significant increase in freezing behavior up to 4 weeks after nerve irradiation. Light microscopic evaluation of nerves removed 7 days post-irradiation, i.e. when maximal allodynia was observed, showed clear evidence of demyelination of large myelinated fibers. These data indicate that photochemically-induced peripheral nerve ischemia is associated with abnormal pain-related behaviors, including mechanical, thermal and cold allodynia and signs of spontaneous pain. The incidence and severity of the behavioral changes are clearly dependent on the exposure time and are probably due to, at least in part, a demyelinaton. These results partly confirm previous data using a similar technique and suggest that this may represent a new animal model for peripheral neuropathy of ischemic origin. The advantages of the present model are its good reproducibility and the fact that the nerve injury can be easily quantified and graded.