The only strategy shown to be consistently beneficial in the treatment of diabetic neuropathy is meticulous control of blood glucose. The largest study of the effects of glycemic control on progression of neuropathy was the Diabetes Control and Complications Trial, which enrolled 1,500 patients. Meticulous control of blood glucose by multiple injections or continuous subcutaneous infusion both delayed the onset of neuropathy and slowed its progression. A weakness of this and other studies of the effect of glycemic control is that they used surrogate measures of improvement (or slowing of progression) of neurologic function. Most used sensory and motor nerve conduction studies and some used vibration perception thresholds. Whether such measures correlate reliably with neuropathy symptom scores, neurologic examination, quality-of-life measures, neuropathic complications (foot ulcers and amputation), and mortality remains controversial. Also, most studies of tight glycemic control do not address the complications of more intensive therapy, among them severe hypoglycemia. Severe hypoglycemia can precipitate acute painful neuropathy, and it markedly increases axonal degeneration in experimental diabetic neuropathy. Finally, all studies have been confined to patients with mild neuropathy; some patients had no clinical evidence of neuropathy. Whether benefit can accrue to patients with more advanced neuropathy is not known. The most physiologic means of achieving glycemic control is through pancreas transplantation; this can result in significant improvement in clinical and electrophysiologic measures of motor and sensory function and slightly improve autonomic function. Strategies to reduce the metabolic consequences of hyperglycemia on nerves and to enhance axonal regeneration are needed to supplement careful glycemic control. Aldose reductase inhibitors hold promise for reducing metabolic nerve injury, but further study is needed. Copyright (C) 1999 Excerpta Medica Inc.