This study was designed to examine the role of endogenous opioid peptide mediation of elevated pain threshold in adult male Sprague-Dawley rats with long-term diabetes mellitus induced by streptozotocin (STZ). Diabetes resulted in a significant elevation in pain threshold as measured by the tail-flick and/or hotplate latency tests. The hypoalgesic response in diabetic rats to hotplate testing developed gradually over a 4-6 week period after a transient hyperalgesia during the first two weeks of diabetes. The elevation of pain threshold achieved peak level by the fourth week after STZ administration, and remained at that level throughout the experimental period (up to 13 weeks). This hypoalgesic state in diabetic animals is thought to be mediated by opioid receptors (i.e. mu and delta). The involvement of the mu receptor is supported by the effect of naltrexone on the STZ-diabetic rats; naltrexone significantly attenuated the increase in tail-flick and hotplate latencies, compared to that of the non-diabetic controls. Furthermore, the concentration of native (free) Met-enkephalin in the spinal cord of STZ-diabetic rats was about 5-fold higher than that of non-diabetic animals. Such high levels of Met-enkephalin suggest the involvement of delta opioid receptors in the hypoalgesic response observed in STZ-diabetic rats. Seven weeks of insulin treatment, initiated after development of the hypoalgesic response, normalized not only plasma glucose level and body weight of diabetic rats, but also returned their antinociceptive latency toward normal. The results of this study showed that long-term diabetes is associated with altered pain threshold and further support the hypothesis for endogenous opioid peptide mediation of hypoalgesia in chronically diabetic rats that can be prevented by insulin treatment.