There is reason to believe that diabetic neuropathy may be related to the accumulation of sorbitol in nerve tissue through an aldose reductase pathway from glucose. Short-term treatment with aldose reductase inhibitors improves nerve conduction in subjects with diabetes, but the effects of long-term treatment on the neuropathologic changes of diabetic neuropathy are unknown. To determine whether more prolonged aldose reductase inhibition reverses the underlying lesions that accompany symptomatic diabetic peripheral polyneuropathy, we performed a randomized, placebo-controlled, double-blind trial of the investigational aldose reductase inhibitor sorbinil (250 mg per day). Sural-nerve biopsy specimens obtained at base line and after one year from 16 diabetic patients with neuropathy were analyzed morphometrically in detail and compared with selected electrophysiologic and clinical indexes. In contrast to patients who received placebo, the 10 sorbinil-treated patients had a decrease of 41.8 +/- 8.0 percent in nerve sorbitol content (P less than 0.01) and a 3.8-fold increase in the percentage of regenerating myelinated nerve fibers (P less than 0.001), reflected by a 33 percent increase in the number of myelinated fibers per unit of cross-sectional area of nerve (P = 0.04). They also had quantitative improvement in terms of the degree of paranodal demyelination, segmental demyelination, and myelin wrinkling. The increase in the number of fibers was accompanied by electrophysiologic and clinical evidence of improved nerve function. We conclude that sorbinil, as a metabolic intervention targeted against a specific biochemical consequence of hyperglycemia, can improve the neuropathologic lesions of diabetic neuropathy.