DNA-based typing has contributed to the understanding of M. tuberculosis epidemiology and evolution. IS6110 RFLP was the first method described and has been used in many epidemiologic investigations. Technological difficulties have hampered the widespread establishment of this method, and it has been found to be of little use in evolutionary studies. PCR-based methods such as spoligotyping and variable number tandem repeat (VNTR) analysis largely overcome these difficulties. Spoligotyping alone is of limited value in epidemiologic investigations due to low discrimination but can be useful in evolutionary studies. Panels of VNTR loci selected from the 59 polymorphic VNTRs described to date have been shown to be useful in both epidemiologic and evolutionary studies. A VNTR type is identified by, first, amplifying a series of PCR fragments each encompassing a different VNTR locus and, second, determining the PCR fragment sizes from which the number of repeats present is calculated. The repeat number present at a series of loci is used as numerical code to describe a type. This chapter describes a high-throughput automated method for VNTR analysis at 15 loci using a capillary fragment analyzer and a manual method using agarose gel analysis.