The clinical syndrome known as kernicterus 1 can usually be prevented even though all the conditions predisposing to it are still unknown. In full-term and premature erythroblastotic infants, exchange transfusion therapy for the prevention of kernicterus has been highly successful. In the nonerythroblastotic premature infant, however, the prevention of kernicterus remains a far more complex and, as yet, incompletely solved therapeutic problem.As in erythroblastotic full-term infants, it appears that some relationship exists in premature infants between the concentration of bilirubin in the plasma or serum and the development of kernicterus. The nature of this relationship, from our experience, is not linear. Nevertheless, exchange transfusions are being performed on premature infants to prevent kernicterus, the same criterion being applied for transfusion as has been used in transfusing erythroblastotic infants, namely, attainment of 20 mg. per 100 ml. concentration of serum total bilirubin.2 In Meyer's study,3 based on observation of the serum concentration of bilirubin in 99 premature infants, the con¬ clusion was expressed that kernicterus is very likely to occur in any infant whose serum (indirect) bilirubin level rises above 18 mg. per 100 ml.Several factors other than erythroblasto¬ sis and infection (bacterial, spirochetal, protozoal, and viral) influence the concen¬ tration of bilirubin in the serum of newborn infants. Vitamin analogues, given in sufficient quantity to either infant4,5 or mother in labor,6 may increase markedly the degree of bilirubinemia and increase the incidence of kernicterus also. Infants with respiratory distress in the first hours or days of life have been observed to have higher concentrations of serum bilirubin than are found in infants of comparable birth weight who do not develop respiratory distress.7 A causal relationship between neonatal hypoxia and kernicterus through higher serum bilirubin concentrations is still to be demonstrated, however. Birth weight,