Tuberculosis is one of the most important infectious threats to human health today with more than a third of the population infected. The organism has co-evolved with humans for more than 10 millennia with periodic epidemics. Throughout this time the organism has undergone deletions of significant parts of its genome allowing us to plot the divergence of important sub-species such as M. africanum and M. bovis. It may be that, as in the case of other pathogens, this has increased the pathogenicity of the organism. Moreover, the impact of other pathogens, notably the human immunodeficiency virus, has resulted in an upsurge in the number and seriousness of cases as it increases the likelihood of reactivation of latent disease. Recent advances in molecular techniques such as insertion sequence, mycobacterial intergenic repeat unit (MIRU) and spoligo-typing have enabled us to plot the evolution of M. tuberculosis over shorter periods. Using these techniques lineages of related organisms can be identified and followed in human populations. Thus we are able to identify lineages with enhanced pathogenicity and the ability to develop antibiotic resistance. The recent description of extensively resistant M. tuberculosis (XDRTB) has once again thrown the spotlight on the importance of antibiotic resistance. M. tuberculosis becomes resistant to antibiotics due to point mutations in the chromosome. For each of the different antibiotics mutations have been described. Rifampicin resistance provides a model system which demonstrates that each mutation is associated with a different reduction in the Darwinian fitness of the organism. The situation becomes more complex when multiple resistance determinants are considered. There is evidence, that, as strains are transmitted between individuals that the fitness deficit is ameliorated as the organism adapts to the new host. This provides an intriguing insight on micro-evolution over very short time frames.