Clinical trials of vaccines against Mycobacterium tuberculosis are well under way and results are starting to come in. Some of these results are not so encouraging, as exemplified by the latest Aeras-422 and MVA85A trials. Other than empirically determining whether a vaccine reduces the number of cases of active tuberculosis, which is a daunting prospect given the chronic nature of the disease, we have no way of assessing vaccine efficacy. Therefore, investigators seek to identify biomarkers that predict vaccine efficacy. Historically, focus has been on the production of interferon-γ by CD4(+) T cells, but this has not been a useful correlate of vaccine-induced protection. In this Opinion article, we discuss recent advances in our understanding of the immune control of M. tuberculosis and how this knowledge could be used for vaccine design and evaluation.