Research into and identification of Mycobacterium tuberculosis can take on a number of facets, many of which involve the use of DNA at one stage or another. The quality and quantity of DNA required will depend on the end-use requirement. For example, good yields of pure, high-molecular-weight DNA uncontaminated by DNA from other sources (i.e., homogeneous) are optimal for the generation of cosmid libraries and sequencing (1), Southern hybridization (2–6), or microarray analysis (7) for genome studies, whereas relatively crude DNA (fragmented DNA or DNA from multiple sources [i.e., heterogeneous]) may be adequate for PCR-based diagnosis (8–12) or amplification of regions of the genome for other purposes, e.g., identification of mutations conferring drug resistance (13,14).