Identifying the correlates of immune protection against HIV-1 acquisition may be an important step in the development of an HIV vaccine, as discussed in a timely commentary by Dr Moog [1]. We were honored to see our work [2] cited in this commentary, but hope to clarify one issue that may cause confusion. Specifically, we did not identify ‘HIV-specific immunoglobulin A (IgA)’ – at least as defined by binding assays such as enzyme-linked immunosorbent assay (ELISA) or western blot – to be a correlate of protection in our prospective study examining the immune correlates of HIV protection. Previous studies of highly exposed, persistently HIV-1 seronegative (HEPS) individuals have described a number of genetic and immune correlates of protection, including innate and adaptive immunity in both the systemic and genital tract compartments. IgA is the predominant antibody isotype at most mucosal surfaces, albeit not that of the lower genital tract [3]. This antibody isotype has been associated with the HEPS phenotype in two distinct capacities. First, ‘HIV-specific IgA’, with HIV specificity defined by binding to various viral antigens in an ELISA or western blot system, has been found in the blood and genital secretions of individuals from some HEPS cohorts [4–10] but not others [11,12]; the latter discrepancies may relate to inter-cohort differences in HIV exposure levels [13] or to variability between different laboratories and different ELISA systems in the detection of low-concentration IgA antibodies [14]. Second, IgA antibodies with HIV inhibitory capacity (most commonly the ability to neutralize HIV replication in peripheral blood mononuclear cells (PBMCs), or to inhibit epithelial transcytosis) have been described in similar HEPS cohorts [4,15,16]. The distinction between ‘HIV-specific IgA’ and IgA antibodies with HIV inhibitory capacity may be an important one to make. Even well defined HIV-binding monoclonal antibodies may lack some HIV inhibitory functions, including the ability to block transcytosis [17]. On the contrary, antibody-mediated inhibition of HIV replication may not require specificity for HIV antigens, as it can be achieved through alternate mechanisms such as the downregulation or blocking of the chemokine (C-C motif) receptor 5 coreceptor on target cells [18], blocking of Dendritic Cell-Specific ICAM-3 Grabbing Nonintegrin [19], or alloimmune responses directed against human leukocyte antigens [20,21]. Although some HEPS studies have described IgA that is both HIV-specific and has anti-HIV functional capacity [4,9], most have focused on either one aspect or the other. In our recent prospective study [2], we purified IgA from the genital secretions of high-risk, HIV-uninfected female sex workers at a single time point, and then followed participants for an average of just over 2 years. IgA neutralization of primary HIV isolates was tested in an operator-blinded fashion, using previously described methods and an assay cut-off that had been validated in HIV low-risk individuals from the region [16]. A PBMC-based assay was used to ensure the detection of IgA reactivity to both HIV protein structures and other possible epitopes involved in HIV fusion and entry. The ability of genital IgA to neutralize HIV was independently correlated with subsequent protection against HIV acquisition (as was HIV-specific proliferation of blood CD8+ T lymphocytes). We did not screen for HIV-specific IgA by ELISA or western blot in a comprehensive fashion, nor did we detect such antibodies in the subset in which these assays were performed (Hirbod T and Broliden K; unpublished data). The mechanism of HIV neutralization was not defined and will be an important area for future research; if HIV-specific IgA was involved, then such work should include mapping the epitope specificity of these antibodies. An additional caveat should be highlighted. Although our prospective study was rigorously performed, with operator-blinded performance of predefined IgA neutralization assays, this was an observational study and shares the limitations of all such studies. Specifically, the direction of causality in the observed association between immune parameters and reduced HIV acquisition cannot be proven. It remains possible that HIV protection was mediated through another mechanism – be that genetic, immune, behavioral or other – and that the immune responses we measured were induced as a result of HIV exposure without infection, rather than mediating the phenomenon. In summary, our recent data demonstrates that IgA antibodies with HIV-neutralizing capacity were present in the lower genital tract of some HEPS female sex workers, and were associated with reduced rates of HIV acquisition. It may be important to make the distinction between IgA antibodies with functional activity against HIV, and ‘HIV-specific’ IgA antibodies (as defined by ELISA). It was the former that we observed to be associated with HIV protection in our prospective study.