BOX, in the presence of O2, selectively destroys bilirubin and its oxidation products. During feedings, active enzyme is recovered from the stools (near 100% recovery in infant rats) and stool bilirubin and bilirubin glucuronides are markedly decreased. Since 1995, on an IND permit, after informed consent, BOX, in PO4 buffer, has been given to eight newborn infants at risk of Ex Tx because of rapidly rising serum bilirubin (SB) in spite of intensive photoRx (BW 1997 to 3960g, GA 32 to 40 wk, 60 to 120 enzyme units/kg/day in one or two hourly feedings or constant OG infusion). Oral BOX was uniformly well tolerated, with no signs of toxicity. Three of the eight were term infants and otherwise healthy; one with severe BO disease (to be described elsewhere), and two with G6PD deficiency with mild hemolysis. Hepatic bilirubin excretion was quite limited in one of these latter infants whose SB rose to 27 mg/dl (BA ratio 6.3mg/g) in spite of intensive photoRx plus BOX, and an Ex Tx was necessary. The remaining five jaundiced BOX-Rx infants had other complications. One was an IDM born at 35/36 wk GA with polycythemia, hypoglycemia, somewhat increased hemolysis (day zero Hct 68%, day 5 Hct 44%) and a rapidly rising SB. After starting oral BOX, SB rise moderated, plateaued and then fell sharply. No Ex was necessary. The four other infants, one term and three preterm, were critically ill, two with severe birth asphyxia, shock and large subgaleal/intracerebral bleeds, one with symtomatic sepsis and poor albumin bilirubin binding affinity (36 wk GA), and one (32 wk GA twin) with severe bruising, probable sepsis and respirtory distress. Only the two infants with large intracranial bleeds required Ex. No Ex was needed in the two other sick infants. A significant increase in bilirubin clearance is suggested by the more rapid than expected fall in SB. We conclude BOX Rx appears promising, completely non toxic, and deserves study in randomized clinical trials. (BOX supplied by Amano International Enzyme Co., Ltd.)}