Tuberculosis (TB) remains a major challenge in public health worldwide. Until today, the only widely used and approved vaccine is the Mycobacterium bovis bacille Calmette-Guerin (BCG). This vaccine provides a highly variable level of protection against the active, pulmonary form of tuberculosis, and practically none against the latent form of TB infection. This disparity in protection has been extensively studied, and for this reason, several groups have focused their research on the quest for attenuated vaccines based on M. tuberculosis or on the identification of latency-associated antigens that can be incorporated into modified BCG, or that can be used as adjuvanted subunit vaccines. In order to seek new potential antigens relevant for infection, some researchers have performed experiments with highly sensitive techniques such as transcriptomic and proteomic analyses using sputum samples from humans or by using mouse models resembling several aspects of TB. In this review, we focus on reports of new mouse models or mycobacterial antigens recently tested for developing vaccine candidates against chronic/latent tuberculosis and its reactivation.