OBJECTIVE - To test the hypothesis that microalbuminuria may show an independent statistical association with diabetic neuropathy. RESEARCH DESIGN AND METHODS - An observational study of a prospectively identified cohort was conducted at the University Medical Center. The cohort consisted of 78 consecutive diabetic patients who fulfilled the criteria of having diabetes for > 10 yr, a normal serum creatinine, urine negative for macroalbuminuria by a commonly used dipstick method, a blood glucose <13.8 mM (<250 mg/dl), and an HbA1 < 11% (normal range 5.5-8.5%). Medical record review established the presence of chronic complications of diabetes. Urine albumin level was measured by radioimmunoassay. Albumin concn > 15 mg/L was used as a cutoff value for microalbuminuria RESULTS - Twenty-five of 78 patients (32%) showed microalbuminuria. Of these, 51% had neuropathy, 39% had retinopathy, 35% arterial hypertension, 17% peripheral vascular disease, and 15% ischemic heart disease. After adjusting for age, sex, and type and duration of diabetes, diabetic neuropathy and hypertension showed a significant association with microalbuminuria. After adjusting for other diabetic complications, diabetic neuropathy showed a significant association with microalbuminuria. CONCLUSIONS - Microalbuminuria is independently associated with diabetic neuropathy. This association lends support to the theory of a vascular etiology for diabetic distal symmetrical neuropathy.