Regardless of the NTHi strain challenged with, groups intranasally immunized with NTHi OMVs showed significant reductions in their nasopharyngeal Gomisin-D colonization rates compared to the nonvaccinated control group, ranging from 400 to 3,900-fold. To our knowledge, this is the first mouse study reporting such high reduction rates for nasopharyngeal colonization studies with NTHi after immunization with a potential NTHi vaccine. We analyzed the protective 4-(Benzyloxy)phenol immune response in more detail and performed two control immunization experiments. We changed the administration route and intraperitoneally immunized mice with IM-1 to avoid induction of a IgA response. These intraperitoneally immunized group exhibited a significant reduction of the nasopharyngeal colonization, but the median colonization rate in intraperitoneally immunized mice was still 100-fold higher compared to the group intranasally immunized with NTHi OMVs. Thus, the best protection upon immunization with NTHi OMVs was observed via the mucosal administration route. Although we cannot exclude other mechanisms, it would be concordant with epidemiological studies on NTHi infections that the IgA response, especially the levels of secretory IgA, are the most important antibodies for the protection against NTHi infections. Besides the induction of a robust antibody response, intranasal immunizations with OMVs will most likely also stimulate antibody-independent mechanism of the immune system, i.e. inflammation, cytokine secretion and recruitment of phagocytes. To investigate the potential role of such general, unspecific responses of the innate immune system we performed intranasal immunizations using OMVs from V. cholerae, which is only distantly related to NTHi. Consequently, these mice induced a robust antibody response against V. cholerae, but not against NTHi. Since V. cholerae OMV immunized mice failed to show protection against challenge with NTHi, nonspecific mechanisms of the immune system cannot account for the reduced colonization observed upon immunization with NTHi OMVs. Consequently, the protective immune response mainly depends on the use of NTHi OMVs and is therefore specific to the OMVs present in the vaccine candidate. Recently, an antibody-independent, CD4 + dependent protective immunity against pneumococcal colonization was described after mucosal immunization using a pneumococcal whole cell killed vaccine. At the current stage similar effects, which are antibody-independent, but specific to NTHi OMVs, cannot be excluded for our model and future investigations using for example antibody-deficient mice will be necessary to investigate all facets of the protective immunity in more detail. As mentioned above, IM-2 contained OMVs from NTHI strains allocated into the LOS groups I, II and III. Consequently, these mice could have raised immunoglobulins directed against the LOS structures of the NTHI strains 2019-R and 3198-R used for the challenge. Thus, we cannot exclude that the observed protection for the IM-2 immunization group relies on anti-LOS antibodies. However, this does not explain the robust induced immune response against heterologous NTHi strains and protection against NTHI strain 3198-R of mice immunized with IM-1, containing solely OMVs derived from NTHI strain 2019-R. Additionally, a recent study by Hirano et al. investigated the protective immune response upon immunization with a LOSbased conjugate vaccine candidate in the mouse model. Immunized mice showed only a 2- to 4-fold reduction in nasopharyngeal colonization after challenge with homologous or heterologous NTHi strains compared to the control group. The higher reduction rates observed in the present study suggest that the protective immune response of the OMV-based vaccine candidate does not depend on antibodies against the LOS structures.