In this study, a metabolomics research approach was used to identify new tuberculosis (TB) markers from sputum, in an attempt to better characterise the disease as well as the metabolic response of the host to Mycobacterium tuberculosis infection. After GCxGC-TOFMS analyses, various multivariate and univariate statistical methods were implemented to identify those compounds best describing the variation between the TB-positive and TB-negative patient groups. The interpretation of these new metabolite markers led to a number of new hypotheses, including: 1) support of the previously proposed citramalate cycle in M. tuberculosis; 2) the interaction of this cycle with an up-regulated glyoxylate cycle during pulmonary M. tuberculosis infection; 3) the increased utilisation of fatty acids and glutamate as alternative carbon sources by M. tuberculosis during pulmonary infection; 4) an alternative mechanism by which the host produces hydrogen peroxide via glucose oxidation, in order to eliminate the bacterial infection; 5) inhibition of the ETC due to pronounced oxidative stress during an active TB disease state, resulting in increased concentrations of various neurotransmitters and other metabolites previously associated with an inborn error of metabolism (MADD/GA type II); and 6) elevated concentrations of neurotransmitters associated with a number of previously described symptoms of TB.