In a double-blind placebo-controlled trial the effect of Sorbinil (250 mg daily) on diabetic neuropathy was examined. After a 2-month run-in placebo period (with three major assessments) 21 patients were randomized to Sorbinil and 10 to placebo, and all were studied for a further 12 months with neurophysiological measurements at 3-month intervals of nerve conduction velocity in multiple nerves, autonomic function tests, vibration thresholds as well as clinical examination and an extensive self-assessment of symptoms. Two subjects on Sorbinil treatment developed a hypersensitivity reaction and were withdrawn. Metabolic control and severity of neuropathy was not significantly different between groups. There were no changes in symptoms as judged by self-assessment scores. No patient entered the trial with neuropathic ulcers but ulceration developed in 4 patients during Sorbinil treatment and in 1 of the placebo group. No beneficial effect of Sorbinil was demonstrated on either the clinical manifestation or on the neurophysiological measurements made in these neuropathic diabetic patients over 12 months of treatment.