Very substantial efforts have been made over the past decade or more to develop vaccines against tuberculosis. Historically, this began with a view to replace the current vaccine, Bacillus Calmette Guérin (BCG), but more recently most candidates are either new forms of this bacillus, or are designed to boost immunity in children given BCG as infants. Good progress is being made, but very few have, as yet, progressed into clinical trials. The leading candidate has advanced to phase IIb efficacy testing, with disappointing results. This article discusses the various types of vaccines, including those designed to be used in a prophylactic setting, either alone or BCG-boosting, true therapeutic (post-exposure) vaccines, and therapeutic vaccines designed to augment chemotherapy. While there is no doubt that progress is still being made, we have a growing awareness of the limitations of our animal model screening processes, further amplified by the fact that we still do not have a clear picture of the immunological responses involved, and the precise type of long-lived immunity that effective new vaccines will need to induce.