Background Previously we have shown that Ag85B-TB10.4 is a highly efficient vaccine against tuberculosis when delivered in a Th1 inducing adjuvant based on cationic liposomes. Another Th1 inducing adjuvant, which has shown a very promising profile in both preclinical and clinical trials, is IC31®. In this study, we examined the potential of Ag85B-TB10.4 delivered in the adjuvant IC31® for the ability to induce protection against infection with Mycobacterium tuberculosis. In addition, we examined if the antigen dose could influence the phenotype of the induced T cells. Methods and Findings We found that vaccination with the combination of Ag85B-TB10.4 and IC31® resulted in high numbers of polyfunctional CD4 T cells co-expressing IL-2, IFN-[gamma] and TNF-[alpha]. This correlated with protection against subsequent challenge with M.tb in the mouse TB model. Importantly, our results also showed that both the vaccine induced T cell response, and the protective efficacy, was highly dependent on the antigen dose. Thus, whereas antigen doses of 5 and 15 [micro]g did not induce significant protection against M.tb, reducing the dose to 0.5 [micro]g selectively increased the number of polyfunctional T cells and induced a strong protection against infection with M.tb. The influence of antigen dose was also observed in the guinea pig model of aerosol infection with M.tb. In this model a 2.5 fold increase in the antigen dose reduced the protection against infection with M.tb to the level observed in non-vaccinated animals. Conclusions/Significance Small changes in the antigen dose can greatly influence the induction of specific T cell subpopulations and the dose is therefore a crucial factor when testing new vaccines. However, the adjuvant IC31® can, with the optimal dose of Ag85B-TB10.4, induce strong protection against Mycobacterium tuberculosis. This vaccine has now entered clinical trials.