The general subject of hyperbilirubinemia of prematurity has been reviewed. Several factors which are now known to affect the concentration of bilirubin in the serum in the neonatal period have been discussed. These are divided into pre- and postnatal factors. It is suggested that the term hyperbilirubinemia should be used in preference to "physiologic jaundice" when referring to infants with concentrations of serum bilirubin of over 10 to 15 mg/100 ml. This roughly separates a group of infants who are potentially endangered and for whose jaundice some specific etiology can frequently be found. Data have been presented which indicate that the incidence of hyperbilirubinemia and kernicterus may vary considerably from one premature nursery to another and within the same unit from time to time. The hypothesis has been offered that this variation may be due to iatrogenic factors. It seems justified to suggest that any premature unit encountering a high incidence of hyperbilirubinemia or kernicterus should carefully examine some of its routine policies. The results of a survey by the author indicate that there is a wide divergence of opinion as to whether hyperbilirubinemia of prematurity should be treated. Considerable variation in the criteria used to select infants for treatment also exist. The only proven effective means of therapy at the present time is exchange transfusion. This has been demonstrated not to be excessively dangerous for premature infants, and available evidence suggests that it is effective in preventing brain damage. Fully realizing the many deficiencies inherent in the concept of a "critical level" of serum bilirubin, the author recommends choosing a concentration of serum bilirubin between 18 and 25 mg/100 ml as the main criterion for the selection of patients for treatment. Some of the newer methods aimed at controlling hyperbilirubinemia are reviewed. These are still in the highly experimental stage and cannot as yet be adequately evaluated.