Molecular strain typing of Mycobacterium tuberculosis has been possible for only about 20 years; it has significantly improved our understanding of the evolution and epidemiology of Mycobacterium tuberculosis and tuberculosis disease. Mycobacterial interspersed repetitive unit typing, based on 24 variable number tandem repeat unit loci, is highly discriminatory, relatively easy to perform and interpret and is currently the most widely used molecular typing system for tuberculosis surveillance. Nevertheless, clusters identified by mycobacterial interspersed repetitive unit typing sometimes cannot be confirmed or adequately defined by contact tracing and additional methods are needed. Recently, whole genome sequencing has been used to identify single nucleotide polymorphisms and other mutations, between genotypically indistinguishable isolates from the same cluster, to more accurately trace transmission pathways. Rapidly increasing speed and quality and reduced costs will soon make large scale whole genome sequencing feasible, combined with the use of sophisticated bioinformatics tools, for epidemiological surveillance of tuberculosis.