Objective: To evaluate the changes in the endoplasmic reticulum stress in the spinal cord of rats with diabetic neuropathic pain (DNP). Methods: Clean-grade healthy male Sprague-Dawley rats, aged 8 weeks, weighing 120-160 g, were fed a high-fat and high-glucose diet for 8 weeks, then diabetes mellitus was induced by intraperitoneal streptozotocin 35 mg/kg and confirmed by blood glucose level >16.7 mmol/L 3 days later.The mechanical paw withdrawal threshold (MWT) and thermal paw withdrawal latency (TWL) were measured at day 14 after injection.The establishment of DNP model was considered successful when MWT and TWL were lower than 85% of the baseline value.Fifteen rats in which the DNP model was successfully established served as DNP group, 15 rats in which the DNP model was not successfully established served as non-NDNP group (NDNP), and another 15 normal rats were selected and served as control group (group C). The MWT and TWL were measured at 3, 7 and 14 days after successful establishment of the model.The rats were then sacrificed, and the lumbar enlargement segments (L_4-6) of the spinal cord were harvested to detect the expression of inositol-requiring enzyme-1α, phosphorylated JNK (p-JNK) and Beclin1 by Western blot. Results: Compared with C and NDNP groups, the MWT was significantly decreased and the TWL was shortened at 3, 7 and 14 days after successful establishment of the model, and the expression of inositol-requiring enzyme-1α, p-JNK and Beclin1 was up-regulated in DNP group (P<0.05). Conclusion The mechanism of development and maintenance of DNP may be related to aggravated endoplasmic reticulum stress in the spinal cord and to autophagy induced by up-regulated expression of p-JNK in rats.