Background:Human intravenous IgG (IVIG) containing specific antibodies protects neonatal rats from septic death. However, IVIG has immunosuppressive properties and clinical trials of IVIG in neonates at risk for sepsis have yielded conflicting results.Hypothesis:This study was designed to test the hypothesis that nonspecific antibodies in IVIG reduce survival in neonatal rats infected withEscherichia coli.Methods:Specific antibodies were adsorbed from IVIG withE colito produce IVIG/anti-E coli−. After transthoracic administration ofE coli, survival was determined in neonatal rats injected intraperitoneally with phosphate-buffered saline, IVIG/anti-E coli− (500 mg/kg) or IVIG containing anti-E coliantibodies (IVIG/anti-E coli+). Complement-mediated hemolytic activity of neonatal rat serum was quantified using sensitized sheep erythrocytes.Results:Compared with placebo, intraperitoneal IVIG/anti-E coli− reduced neonatal survival afterE coliinfection. In contrast, IVIG/anti-E coli+ protected infected animals. Both IVIG/anti-E coli− and IVIG/anti-E coli+ impaired the complement-mediated hemolytic activity of neonatal rat serum.Conclusions:IVIG contained (1) nonspecific antibodies that reduced survival in neonatal rats infected withE coliand (2) protective anti-E coliantibodies that enhanced survival in neonatal rats infected withE coli. We speculate that in clinical trials of IVIG to treat or prevent neonatal sepsis, inconsistent results may be caused, in part, by lot-to-lot variations in the ratio of immunosuppressive, nonspecific antibodies to protective, specific antibodies.