Newborn infants, and especially premature ones, are uniquely susceptible to severe and overwhelming bacterial infections. The reasons for this are not completely understood. A number of defects in the neonate's host defense system have been described, including deficiency of opsonic antibodies, abnormalities in the number and function of polymorphonuclear leukocytes, and depressed levels of the major complement components. Experimental animal studies in our laboratory have indicated that many of these defects can be overcome through the administration of intravenous immune globulin. Recently several investigators have attempted to treat or prevent bacterial infections in human neonates with intravenous immune globulin. We have found that doses as high as 750 mg/kg can be administered to premature infants without detectable side effects. This results in IgG levels comparable with those in term infants and adults. Limited studies in the literature suggest that intravenous immune globulin therapy may decrease morbidity and mortality from infection in premature human infants. Furthermore, prophylactic treatment of preterm infants may serve to prevent the development of infection in some instances. Optimal management of preterm and some term human infants may involve immunotherapy with intravenous immune globulin.