Résumé : BACKGROUND: There is a need for new vaccines for tuberculosis (TB) that protect against adult pulmonary disease in regions where BCG is not effective. However, BCG could remain integral to TB control programmes because neonatal BCG protects against disseminated forms of childhood TB and many new vaccines rely on BCG to prime immunity or are recombinant strains of BCG. Interferon-gamma (IFN-gamma) is required for immunity to mycobacteria and used as a marker of immunity when new vaccines are tested. Although BCG is widely given to neonates IFN-gamma responses to BCG in this age group are poorly described. Characterisation of IFN-gamma responses to BCG is required for interpretation of vaccine immunogenicity study data where BCG is part of the vaccination strategy. METHODOLOGY/PRINCIPAL FINDINGS: 236 healthy Gambian babies were vaccinated with M. bovis BCG at birth. IFN-gamma, interleukin (IL)-5 and IL-13 responses to purified protein derivative (PPD), killed Mycobacterium tuberculosis (KMTB), M. tuberculosis short term culture filtrate (STCF) and M. bovis BCG antigen 85 complex (Ag85) were measured in a whole blood assay two months after vaccination. Cytokine responses varied up to 10 log-fold within this population. The majority of infants (89-98% depending on the antigen) made IFN-gamma responses and there was significant correlation between IFN-gamma responses to the different mycobacterial antigens (Spearman's coefficient ranged from 0.340 to 0.675, p = 10(-6)-10(-22)). IL-13 and IL-5 responses were generally low and there were more non-responders (33-75%) for these cytokines. Nonetheless, significant correlations were observed for IL-13 and IL-5 responses to different mycobacterial antigens CONCLUSIONS/SIGNIFICANCE: Cytokine responses to mycobacterial antigens in BCG-vaccinated infants are heterogeneous and there is significant inter-individual variation. Further studies in large populations of infants are required to identify the factors that determine variation in IFN-gamma responses.