The reported prevalence of diabetic polyneuropathy varies from 5 to 80%. This unsatisfactory state may relate to evaluation of different patient groups, different minimal criteria for the diagnosis of neuropathy, and different degrees of surveillance. To made matters worse, patients with polyneuropathy tend to be equated ignoring differences in severity. To remedy this situation, four recommendations are made: (1) population-based patients should be studied, (2) nerve conduction should be used to set minimal criteria for neuropathy because the test is objective, sensitive, and repeatable, (3) validated tests of symptoms and deficits should also be used because clinical manifestations of neuropathy cannot be accurately inferred from electrophysiologic measurements, and (4) approaches to staging severity of neuropathy should be developed and used in expressing abnormality. To this end minimal criteria for the diagnosis of diabetic polyneuropathy have been proposed, and validated tests to assess neuropathic symptoms and sensory deficits have been developed. In this report we also propose a staging approach utilizing nerve conduction and neurologic history and examination and validated tests of neuropathic symptoms and deficits.