Pain threshold was assessed via tail flick latency in alloxan-diabetic, streptozotocin-diabetic, BB/W-diabetic, and control laboratory rats. In addition, tail flick latency was determined under conditions of both euglycemia (60-120 mg/dl) and hyperglycemia (greater than 250 mg/dl). Conditions of hyperglycemia resulted in a significant decrease in tail flick latency in both diabetic and control animals. More interestingly, tail flick latency was returned to control values in diabetic rats following normalization of blood glucose levels. It is concluded that elevated blood glucose levels result in a decrease in pain threshold in both diabetic and control rats. In addition, a permanent state of painful symptoms may be avoided in clinical diabetes by improvement of diabetic control.