Currently available human immunoglobulin preparations for intravenous use (IVIGs) are being used (with antibiotics) by some physicians for therapy of sepsis in newborns. Most neonatal sepsis and/or meningitis in this country is caused by group B Streptococcus (GBS), and most of these cases are due to type III GBS (III-GBS). The killing of III-GBS in vitro is dependent on specific IgG antibody. Adequate serum levels of specific III-GBS antibody protect the exposed newborn from the development of invasive disease. Therefore, III-GBS was used as a model to evaluate the activity of three IVIG preparations available for clinical use. Specific antibody levels, in vitro opsonophagocytic killing, and protective efficacy in animal models revealed differences in activity for III-GBS between the three IVIG preparations as well as between IVIG lots from the same manufacturer. Furthermore, it was found that the effect of IVIG using one of the assay methods may not reliably predict activity obtained using the other assays. These data document the inability to predict functional activity against a specific pathogen such as GBS on the part of a lot of IVIG chosen at random. In view of these findings and of the limited data evaluating clinical efficacy, IVIG cannot be recommended at this time for use in the therapy of infectious diseases such as neonatal sepsis.