To access the application of spacer oligotyping (Spoligotyping) and Multiple Locus VNTR(MLVA) in epidemiological studies of Mycobacterium tuberculosis.224 clinical isolates of M. tuberculosis were collected and typed by Spoligotyping and MLVA respectively, to compare the results of both methods and to access their application in epidemiological studies of M. tuberculosis.Data from Spoligotyping showed that 224 strains presented 55 kinds of genotypes. Of these, 39 were represented by a unique isolate, with the remaining 185 isolates being grouped in 16 clusters whereas the result of MLVA showed that 224 strains presenting 160 kinds of genotypes. Of these, 132 were represented by a unique isolate, with the remaining 92 isolates being grouped in 28 groups. Data from the combination of Spoligotyping and VNTR showed that 224 strains presenting 179 kinds of genotypes. Of these, 159 were represented by a unique isolate, with the remaining 65 isolates being grouped in 20 groups. There was significant difference noticed among M. tuberculosis between Hunan and Anhui in the proportion of Beijing family (P < 0.001). The proportion of Beijing family in Anhui was higher than that in Hunan.Results from this direct comparison studies demonstrated that MLVA analysis was more effective than Spoligotyping in discriminating individual M. tuberculosis isolates. However, Spoligotyping had an advantage over MLVA in identifying Beijing family strains and M. bovis. Taking Spoligotyping as a first-line typing technique and VNTR as second-line typing technique, the arrangement would improve the effectiveness of epidemiological investigation and pathological inspection of tuberculosis. The strains in different regions seemed to have had different characteristics.