An important role for the immune system is to maintain protective immunological memory to a wide variety of pathogens encountered over one's lifetime, while still leaving the host able to respond to newly encountered pathogens. Vezys et al. make the interesting observation that it is possible to repeatedly immunize mice in ways that allow for development of high numbers of memory CD8 T cells without depleting pre-existing memory cells specific for other pathogens. This study, which offers promise in developing potent vaccination schemes, is seemingly at odds with work published by us in the 1990s showing a loss in CD8 memory cells after a series of infections. In their reply, Vezys et al. mention that we may have misinterpreted our data because we reported the putative loss of memory T cells as per cent rather than total number, but here we represent the data in those studies as total cell number. We show here in Fig. 1 that a series of infections can indeed reduce the total number of memory cells, indicating that vaccination strategies need to consider this issue.