Summary The present study reports an analysis of 500 cases of juvenile diabetes(onset of the disease during the first two decades of life) in a total series of 6,000 diabetics observed during a period of twenty-seven years. Thus the incidence of diabetes in patients 20 years of age or less was 8.3 per cent. The sex distribution in the 500 cases of diabetes in children was practically equal, with 255 (51 per cent) boys and 245 (49 per cent) girls. Heredity and infection are evidently the prime etiologic factors in thedevelopment of juvenile diabetes. In this series there was an hereditary or familial history of diabetes in 31.5 per cent, with the incidence in forty-six Jewish children of the group somewhat higher, i. e., 35.5 per cent. Infections, principally the acute exanthemata, play a considerable role in precipitating the onset of diabetes. In this group of 500 cases, there was a history of antecedent infection in 164. In the majority of these, the diabetes appeared within two months after the infection. A comparison of glucose tolerance tests in adults and children and of thehistories of diabetic patients in the two age groups shows the increasing importance of obesity, endocrine disturbances, and degenerative changes, particularly arteriosclerosis, in the causation of disturbances in carbohydrate metabolism and in the development of diabetes, with advancing age. Of this series of 500 juvenile diabetics, 364 were traced to the present time. Of these, 303, or 83.2 per cent are living, and sixty-one (16.7 per cent) are dead. Of the 303 living patients, 41.25 per cent have been followed for ten years or less since the onset of the diabetes, 43.23 per cent have been observed from ten to twenty years, and 15.51 per cent have had the disease more than twenty years. Diabetic coma was the principal cause of death in this series, accountingfor thirty-one of the sixty-one deaths. Only six of these thirty-one patients were under my immediate care at the time of death; of these six, one was treated in the preinsulin era and three were moribund and died within one hour after admission to the hospital. The others were under the care of their family physicians, mostly in small towns, at the time of the emergency. The second most important cause of death was intercapillary glomerulosclerosis, which brought death to eleven of the sixty-one patients. The other deaths occurred after infections, surgical operations, or accidents. Deaths in these sixty-one patients occurred from within less than a year to twenty-four years after onset of the diabetes. The largest number died within the first year, i. e., eighteen of the sixty-one (29.5 per cent). Thirty-seven of the deaths occurred in males and twenty-four in females. As shown by the mortality figures, diabetic coma is the most serious complicationthreatening the child with diabetes. According to the records in this series, as experience has been gained by physicians and as the training of younger physicians in the handling of emergencies has improved, the incidence of diabetic coma has become considerably less than in the first decade after insulin was discovered. In this series, there were 218 patients seen from 1920 to 1934, and of these, fifty are known to have had diabetic coma (23 per cent). In the period 1934 to 1948, twenty-seven patients (9.5 per cent) had diabetic coma. Of the seventy-seven patients with this complication, fifty-four were under my personal care at the time of the emergency, with a total of six deaths. Twenty-three patients received treatment for the diabetic coma by their family physicians, and in this group there were eighteen deaths. The optimism concerning the fate of diabetic children which representedthe general concensus in the years immediately following the discovery of insulin has been modified considerably in recent years by the high incidence of premature arteriosclerotic changes in these patients, leading to retinopathy and glomerular changes in the kidney which have led to blindness and early death in many instances. In this series, eleven of the sixty-one deaths (18 per cent) were due to intercapillary glomerulosclerosis. In a series of eighty-three patients in this series whom I have been able to observe by regular periodic examinations over a period of years, twenty-six (31.3 per cent) show albuminuria. In this series, the incidence of albuminuria is not significantly higher in patients who have had diabetes for ten to twenty-five years than in the group who have had the disease less than ten years. Other authors report an increasing incidence with longer duration of diabetes.