In previous studies, we have described a live recombinant BCG vaccine (rBCG30) overexpressing the 30 kDa major secretory protein of Mycobacterium tuberculosis that induces greater protective immunity against tuberculosis than the current vaccine in the demanding guinea pig model of pulmonary tuberculosis. In this study, we have investigated the impact of vaccine dose on the development of cell-mediated and protective immunity in the guinea pig model. We found that the protective efficacy against M. tuberculosis aerosol challenge of both BCG and rBCG30 was essentially dose-independent over a dose range of 10(1)-10(6) live organisms. As previously observed, rBCG30 was more potent, reducing colony-forming units (CFU) below the level observed in animals immunized with the parental BCG vaccine by 0.7 logs in the lungs and 1.0 logs in the spleen (P<0.0001). To gain a better understanding of the influence of dose on bacterial clearance and immunity, we assessed animals immunized with 10(1), 10(3), or 10(6)CFU of rBCG30. The higher the dose, the higher the peak CFU level achieved in animal organs. However, whereas humoral immune responses to the 30 kDa protein reflected the disparate CFU levels, cell-mediated immune responses did not; high and low doses of rBCG30 ultimately induced comparable peak lymphocyte proliferative responses and cutaneous delayed-type hypersensitivity responses to the 30 kDa protein. We estimate that the amount of the 30 kDa protein required to induce a strong cell-mediated immune response when delivered via 10 rBCG30 organisms is about 9 orders of magnitude less than that required when the protein is delivered in a conventional protein/adjuvant vaccine. This study demonstrates that a very low inoculum of rBCG30 organisms has the capacity to induce strong protective immunity against tuberculosis and that rBCG30 is an extremely potent delivery system for mycobacterial antigens.