Risk factors associated with diabetic microvascular complications, with special reference to ethnic origin, were looked for in 231 young Jewish insulin-dependent diabetes mellitus (IDDM) patients with duration of diabetes ≥10 yr. Median age at diagnosis of diabetes was 9.2 yr (range 0.04-26.2 yr), and median duration of the disease was 15.3 yr (range 10.0-37.2 yr). Sixty-three percent of the patients were Ashkenazi Jews, and 37% were non-Ashkenazi Jews. HbA1 was evaluated every 3 mo in the last 10 yr of follow-up, and albumin excretion rate was tested in three 24-h urine collections. Direct and indirect ophthalmoscopy was performed every year since diagnosis of diabetes, and if retinal pathology was suspected, color photographs were taken. Microalbuminuria was detected in 31% and macroalbuminuria in 7% of the patients. Nonproliferative and proliferative retinopathy was found in 44 and 12% of the patients, respectively. On logistic regression analysis, two variables were significantly and independently associated with diabetic nephropathy - non-Ashkenazi origin and mean HbA1 values over the first 5 of 10 yr of follow-up. Variables significantly and independently related to diabetic retinopathy were non-Ashkenazi origin, mean HbA1 values over the last 10 yr of follow-up, and duration of diabetes. Because non-Ashkenazi Jews in Israel are of lower socioeconomic status than Ashkenazi Jews, we stratified our patients according to their socioeconomic parameters, median HbA1 values, and duration of diabetes. Non-Ashkenazi patients were at a higher risk to develop complications in all strata. We further stratified patients into four quartiles according to mean HbA1 values; there was a steep increase in the risk to develop macroalbuminuria in the 4th quartile of HbA1 (odds ratio [OR] 4.3 vs. 1.2 in the 3rd quartile) and proliferative retinopathy (OR 13.0 in the 4th quartile of HbA1 vs. 2.8 in the 3rd quartile). We conclude that non-Ashkenazi Jewish IDDM patients are at significant risk to develop microvascular complications, independent of their glycemic control, duration of diabetes, and socioeconomic status. Careful follow-up and special efforts toward improving glycemic control should be focused on high-risk subgroups of patients.