The efficacy of the conventional bacille Calmette-Guérin (BCG) vaccine is often questioned because of conflicting trial results [1]. In this context, a new BCG vaccine developed by Grode et al. [2] has been seen as a favourable development in tuberculosis (TB) control [3]. BCG vaccine development is a challenge for many reasons, not least of which is the complex immunology of mycobacterial infections. The pathogenesis of TB, and prevention of TB by BCG, is predominantly a function of cell-mediated immunity. Mycobacterium tuberculosis (MTB) and BCG are phagocytosed, but remain viable for some time within the phagosome. There, BCG antigens must be presented to major histocompatibility complex (MHC) class II molecules to activate CD4+ cells, and to MHC class I molecules to activate CD8+ cells [3]. However, both MTB and BCG fail to activate CD8+ cells sufficiently. The relative failure of activation of CD8+ cells is thought to be related to the enzyme urease, which keeps the intracellular pH high and thus inhibits presentation of antigens to T cells which would otherwise activate the CD8+ response. Grode et al. [2] have engineered a BCG strain which overcomes this problem and allows release of antigen into the cell and thereby stimulates a CD8+ cell response in addition to the CD4 response seen with conventional BCG. They did this first by engineering a strain of BCG which secretes listeriolysin (Hly), an enzyme from Listeria monocytogenes which perforates cell membranes.