INTRODUCTION: Mycobacterium tuberculosis kills more people than any other bacterial pathogen. New drugs are required to shorten the treatment time and provide a viable therapy for drug-resistant and latent forms of tuberculosis. The tuberculosis field has advanced considerably since the publication of the M. tuberculosis genome sequence. Today, researchers can build a high definition map of the pathogen's traits and behavior and select individual targets for chemical disruption. AREAS COVERED: This review examines the discovery of current clinical and candidate tuberculosis drugs. It outlines recent developments in the selection of molecular targets for the discovery of new anti-mycobacterial agents. It appraises techniques that incorporate target knowledge into the screening protocol. These techniques include in silico, in vitro enzyme-based, differential antisense sensitivity and gene expression screening systems. The review also looks ahead to further techniques that may be applied in tuberculosis drug discovery. EXPERT OPINION: The adoption of an 'either/or' approach to targeted or random tuberculosis drug screening is not expected. The historical success of random screening in providing the tuberculosis drugs currently in clinical use is likely to ensure that non-targeted protocols retain an important role in drug screening. However, a number of M. tuberculosis inhibitors in lead optimization and preclinical development have been discovered using targeted methods. Realization of the first clinically-approved tuberculosis drugs derived from targeted screening and continued refinements in targeted screening technologies are likely to increase the adoption of targeted approaches in the future.