Newborn infants may have IgG deficiencies that increase their susceptibility to bacterial infection. To determine whether intravenous immune globulin (IVIG) therapy improves survival rates in early-onset sepsis, we prospectively entered 753 neonates (birth weight 500 to 2000 gm, gestation ≤34 weeks, age ≤12 hours) into a multicenter, double-blind, controlled trial. Blood culture specimens were obtained and infants randomly assigned to receive 10 ml (per kilogram) intravenously of a selected IVIG (500 mg/kg) or albumin (5 mg/kg) preparation. Maternal and neonatal risk factors were not different between groups. Thirty-one babies (4.2%) had early-onset sepsis; the causative organisms were group B streptococcus (12 babies), Escherichia coli (6), and others (13). Of these 31 neonates, 7 (23%) died. Total serum IgG was higher for 7 days after IVIG therapy than after albumin treatment (p<0.05). During these 7 days, 5 (30%) of 17 albumin-treated and none of 14 IVIG-treated patients died (p<0.05). The survival rate at 56 days of age, however, was not significantly improved. Group B streptococcus type-specific IgG antibody was significantly increased after IVIG treatment and appeared to be related to the amount of IVIG specific antibody. Infusion-related adverse reactions were less frequent in patients receiving IVIG therapy (0.5%) than in those receiving albumin. The IVIG therapy in neonates with early-onset sepsis, while reducing the early mortality rate, did not significantly affect the overall survival rate. Further studies are necessary to confirm these findings and to determine more effective therapeutic regimens.