We examined the tail-flick response to various heat intensities in diabetic and non-diabetic mice. Heat intensities were set to one of six values by adjusting the source of voltage for a 50-W projection bulb to 20, 25, 35, 50, 65 and 80 V. Tail-flick latencies at source voltages of 35 and 50 V in diabetic mice were significantly shorter than those in non-diabetic mice. However, tail-flick latencies at 25, 65 and 80 V in diabetic mice were not significantly altered. Although tail-flick latencies in non-diabetic mice were not affected by i.t. pre-treatment with CI-988, a selective cholecystokinin B (CCK(B)) receptor antagonist, those at 35 and 50 V in diabetic mice were significantly increased. In non-diabetic mice, i.t. pre-treatment with cholecystokinin octapeptide (CCK-8), at a dose of 0.3 ng, decreased tail-flick latencies at 35 and 50 V. Furthermore, the attenuation of tail-flick latencies induced by i.t. pre-treatment with CCK-8 in non-diabetic mice was reversed by i.t. pre-treatment with CI-988. Protein kinase C (PKC) activator phorbol-12, 13-dibutyrate (PDBu)-induced reduction in the tail-flick latencies at heat intensities of 35 and 50 V in non-diabetic mice was dose-dependently and significantly reversed by i.t. pre-treatment with CI-988. On the other hand, the CCK-8-induced thermal hyperalgesia and allodynia at heat intensities of 35 and 50 V in non-diabetic mice were inhibited when PKC activity was inhibited by i.t. pre-treatment with calphostin C. These results indicate that the thermal allodynia and hyperalgesia in diabetic mice may be due, at least in part, to the activation of CCK(B) receptors followed by the activation of PKC in the spinal cord.