BACKGROUND: There is little information on the risk factors for diabetic polyneuro-pathy other than glycemic control and duration of diabetes. The relation between diabetic polyneuropathy and hypoinsulinemia is controversial. This study is to determine whether hypoinsulinemia is an additional factor influencing the development of polyneuropathy in patients with type 2 diabetes. METHODS: We performed an oral glucose tolerance test with C-peptide measure-ment in 1'P2 patients with type 2 diabetes. Peripheral polyneuropathy was diagnosed when the patients had both typical symptoms of polyneuropathy and abnormal physical findings or NCV. We analysed the relation between metabolic variables including fasting and postprandial C-peptide levels and diabetic polyneuropathy. RESULTS: In addition to retinopathy and nephropathy, duration of diabetes, low C-peptide level (fasting and postprandial), insulin use, and high HDL-cholesterol level were associated with diabetic polyneuropathy. Multivariate logistic regression model revealed that an independent assoclation of diabetes duration and postprandial 2-hour C-peptide concentration with polyneuropathy. When we stratified the patients into the two groups according to the median duration of diabetes (8 years), the association of low postprandial C-peptide concentration with polyneuro-pathy was significant only in the shorter duration group(< 8 years). However, significant association of HbA(1c) level was shown in the longar duration group. CONCLUSION: Decreased insulin secretory function of the pancreas as well as increased duration of diabetes was indepandently associated with diabetic polyneuropathy in patients with type 2 diabetes. Hypoinsulinemia might be an additional risk factor for the development of diabetic polyneuropathy in type 2 diabetic patients, particularly with short duration, To confirm these relationship further longitudinal study in a large cohort is necessary.