One hundred and twenty-nine (87%) of a total county population of 150 eligible diabetic children together with 144 age- and sex-matched control children participated in a longitudinal, epidemiological study of the evolution of diabetic microvascular disease. At enrollment the median (range) age of the diabetic children was 12.5 (3.7-16.8) years with a median diabetes duration of 2.9 (0.1-13.4) years and a median HbAl of 11.1 (6.8-17.9)%. Two sets of measurements were made over a period of 18 months for all indices of microvascular disease, while autonomic function was studied on one occasion. Urinary albumin excretion in diabetic children was assessed from all voidings during two timed 48-h urine collections and was expressed as urinary albumin/creatinine ratios (ACR). Blood pressure (BP) was measured using a random zero sphygmomanometer. Autonomic function was assessed by pupillary adaptation in darkness, using a portable Polaroid pupillometer, and by heart rate (HR) variation recorded by dedicated computer. Vibration sensation thresholds (VST) (as indices of peripheral neuropathy) were recorded using a Biothesiometer. Limited joint mobility (LJM) was assessed by the "prayer sign". Five (3.9%) diabetic children presented raised mean ACR in more than two of four 24-h urine collections. Fourteen (10.8%) diabetic children were identified as having persistently raised BP during both study periods. Impaired HR response in one HR test was observed in 20 (15.5%) diabetic children, while ten (7.7%) diabetic children demonstrated abnormalities in two or more HR tests. Reduced pupillary adaptation in darkness was found in eight (7.9%) diabetic children. Persistent vibration sensation impairment (VST) in lower limbs was detected in eight (6.2%) diabetic children, while LJM was present in 12 (9.3%) diabetic children. Eight of the 129 diabetic children (6.2%) were found to have abnormality in two and one in three indices of microvascular and autonomic function. Six of nine children had coexistence of impaired autonomic neuropathy and nephropathy. These nine children were diagnosed at a younger age than the rest of the diabetic population (5.1 vs 8.0 yr, p=0.002). Four of nine were aged >11 yr and five of nine had had diabetes for >5 yr. Thus, a constellation of microvascular and neurological abnormalities were demonstrable in a small proportion of diabetic children, who were younger than the rest of the population at the time of onset of their disease. Longitudinal study of this population will demonstrate the clinical significance of these findings.