There is great need for better control measures for tuberculosis (TB). High-throughput analyses, such as transcriptomic and metabolic profiling, offer a promising path toward clinically useful biosignatures. With the help of biomarkers, it will be possible not only to reliably perform diagnosis but also to gain a better understanding of the disease process and, in the future, even predict the onset of disease in infected individuals. Biomarkers based on transcriptomic and metabolic profiles as well as on cytokine composition provide important insights into the basic biological principles of TB and give an opportunity to reliably distinguish TB patients from healthy individuals. Use of biomarkers for point-of-care diagnosis, however, is still a distant goal, which to achieve will require extensive analysis of TB biosignatures across different cohorts and a combination of different platforms.