BACKGROUND: Mycobacterium tuberculosis is one of the most clinically significant infectious agents. Especially during mass outbreaks, accurate identification and monitoring are required. The proportion of Beijing family members is very high among infecting strains, and spoligotyping is not suitable for strain typing. Therefore, we studied the homogeneity of isolates using the mycobacterial interspersed repetitive units-variable number of tandem repeats (MIRU-VNTR) method and identified its utility for carrying out molecular epidemiologic analysis. METHODS: Eighty-one clinical M. tuberculosis isolates that had previously been analyzed by spoligotyping were used in this study. We used the 12 standard MIRU loci and further four exact tandem repeat (ETR) loci (ETR-A, -B, -C, and -F). Four strains each of randomly selected Beijing and Beijing-like families were subjected to IS6110- restriction fragment length polymorphism analysis. RESULTS: All 81 samples showed amplification products of all VNTR loci, and all of them showed differences in at least one locus. The calculation of the Hunter-Gaston diversity index (HGDI) for MIRU-VNTR gave the value of 0.965. Discriminatory index in the six loci (MIRU-10, -16, -26, -31, -39, and ETR-F) were found to be highly discriminated (HGDI >0.6). Beijing and Beijing-like family isolates were discriminated into different MIRU-VNTR types. CONCLUSIONS: MIRU-VNTR analysis by using well-selected loci can be useful in discriminating the clinical M. tuberculosis isolates in areas where the Beijing family is predominant.