No diabetes physician would question the fact that problems relating to nerve damage in diabetes mellitus are extremely common and result in much patient morbidity and unhappiness [1]. However, as the cause and natural history of diabetic peripheral neuropathy remain unknown, attempts at defining minimal criteria for the presence of neuropathy have proved difficult [2]. This is probably the most important reason for the widely varied estimates of the prevalence of diabetic peripheral neuropathy. Other factors include differing patient selection criteria, the small size of cohorts and the employment of different diagnostic tests. Some studies have used the presence of neuropathic symptoms [3] alone for the diagnosis of neuropathy while others have used both symptoms and signs of neuropathy as minimal criteria for diagnosis [4,5]. The San Antonio Conference on Diabetic Neuropathy [6] recommended that for full classification of diabetic neuropathy, at least one measure from each of the following categories need Diabetologia (1996) Summary The EURODIAB IDDM Complications Study involved the examination of 3250 randomly selected insulin-dependent diabetic patients, from 31 centres in 16 European countries. Part of the examination included an assessment of neurological function including neuropathic symptoms and physical signs, vibration perception threshold, tests of autonomic function and the prevalence of impotence. The prevalence of diabetic neuropathy across Europe was 28 % with no significant geographical differences. Significant correlations were observed between the presence of diabetic peripheral neuropathy with age (p < 0.05), duration of diabetes (p < 0.001), quality of metabolic control (p < 0.001), height (p < 0.01), the presence of background or proliferative diabetic retinopathy (p < 0.01), cigarette smoking (p < 0.001), high-density lipoprotein cholesterol (p < 0.001) and the presence of cardiovascular disease (p < 0.05), thus confirming previous associations. New associations have been identified from this study -namely with elevated diastolic blood pressure (p < 0.05), the presence of severe ketoacidosis (p < 0.001), an increase in the levels of fasting triglyceride (p < 0.001), and the presence of microalbuminuria (p < 0.01). All the data were adjusted for age, duration of diabetes and HbA 1c . Although alcohol intake correlated with absence of leg reflexes and autonomic dysfunction, there was no overall association of alcohol consumption and neuropathy. The reported problems of impotence were extremely variable between centres, suggesting many cultural and attitudinal differences in the collection of such information in different European countries. In conclusion, this study has identified previously known and new potential risk factors for the development of diabetic peripheral neuropathy. [Diabetologia (1996[Diabetologia ( ) 39: 1377[Diabetologia ( -1384