Tuberculosis remains a major health threat with 9 million new cases and 2 million deaths annually. Although a vaccine (BCG) is available, it only protects against childhood tuberculosis but not against pulmonary tuberculosis of adults, the most prevalent form of this disease. Control of tuberculosis critically depends on design of novel vaccine. Because one third of the world population is infected with M. tuberculosis and therefore at increased risk of developing disease, both pre-exposure and post-exposure vaccines are needed. Currently, several pre-exposure vaccines are in, or have passed, clinical phase I trials. Current vaccine candidates are subunit vaccines aimed at booster of prime with current BCG and recombinant BCG with improved efficacy aimed at substituting BCG. Obviously, both strategies can be combined as prime with recombinant BCG and booster with a subunit vaccine. Despite some success of these candidates in preclinical models, ultimately their efficacy can only be revealed in clinical trials. These are long-term endeavours ranging from 5 to 10 years and require large study populations. Hence, biomarkers are urgently needed which allow prediction of vaccine success or failure at an early time point. Because the vast majority of tuberculosis cases live in developing countries, novel support strategies need to be exploited. These include public-private partnerships as well as support from foundations, governmental and non-governmental organizations. Most vaccines against tuberculosis have been developed in academic research labs. Public-private partnerships are needed to transfer these vaccines from the bench to the field. Thus, industry needs to be attracted, e.g. through tax reductions, tiered-price system and guaranteed purchase of the vaccine. Finally, capacity building is needed to provide the infrastructure for clinical trials in developing countries.