Mycobacteria expressing recombinant antigens are already being developed as vaccines against both infections and tumors. Little is known about how dendritic cells might process such antigens. Two different mycobacterial species, the fast-growing Mycobacterium smegmatis and the slow-growing M. bovis M. bovis BCG, were engineered to express a model tumor antigen, the K(b)-restricted dominant cytotoxic T-lymphocyte epitope OVA(257-264). Recombinant M. bovis BCG but not recombinant M. smegmatis conferred protection to mice challenged with the B16-OVA tumor cell line. We went on to investigate whether the contrast in antitumor efficacy could be due to differences in how dendritic cells process antigen from the two mycobacterial strains for class I presentation. Both strains of mycobacteria caused phenotypic maturation of dendritic cells, but recombinant M. smegmatis infection led to a greater degree of dendritic cell maturation than recombinant M. bovis BCG infection. Antigen from recombinant M. smegmatis was processed and presented as OVA(257-264) on K(b) molecules by the dendritic cell line DC2.4 but not by bone marrow-derived dendritic cells (BMDC) or splenic dendritic cells. In contrast, antigen from recombinant M. bovis BCG was presented by all three dendritic cell types as long as the mycobacteria were viable. Such presentation was dependent on proteasome function and nascent major histocompatibility complex (MHC) class I molecules in DC2.4 cells but independent of the proteasome and transporter associated with antigen processings (TAP) in BMDC and splenic dendritic cells. These data demonstrate for the first time that antigen vectored by the slow-growing M. bovis BCG but not that vectored by fast-growing, readily destroyed M. smegmatis is processed and presented on MHC class I by in vitro-generated dendritic cells, which has implications for recombinant microbial vaccine development.