The study of the normal and pathologic metabolism of premature infants may be regarded as interesting and important from more than one point of view. The literature on the subject is fairly meager, though the knowledge has been increased greatly by recent investigations. My co-workers and I determined to institute further examinations into some aspects of the carbohydrate metabolism of premature children. The direct motive for this was that we wished to see how far two rare disturbances in the carbohydrate metabolism of children, which probably have existed since birth, and which we had recently observed, might be looked on as a continuation of a fetal condition. These disturbances were a case of levulosuria1 and one of chronic marked hypoglycemia without any hypoglycemic reaction;2 the hypoglycemia, combined with acetonuria, was always noted after the patient (a boy, now 8 years old) had been fasting for a short time.