Abstract： Background Mycobacterial interspersed repetitive units-variable number tandem repeat (MIRU-VNTR) and Beijing family typing based on detecting the deletion of RD105 sequence are two common genotyping methods used to study the molecular epidemiologic characteristics of Mycobacterium (M.) tuberculosis.We collected 218 strains of M.tuberculosis between 2004 and 2006 in the Linxia Hui Autonomous Prefecture of Gansu province in Northwest China.Methods MIRU-VNTR analysis and Beijing family typing based on detecting the deletion of RD105 sequence were used to type the 218 strains,and their typing power was evaluated to look for practical and efficient genotyping methods suitable for the region.Results The MIRU typing yielded 115 distinct genotypes,including 98 unique isolates and 17 different clusters containing 120 isolates (55.05％); the cluster rate was 47.25％.By detecting the deletion of RD105 sequence,188 of 218 (86.23％) isolates belonged to Beijing family.Combination of Beijing family typing and MIRU typing yielded 118 distinct patterns,including 101 unique isolates and 17 clusters containing 117 isolates (54.13％).The largest cluster contained 58 strains with MIRU genotype of 223325173533 which contained 50 strains belonging to Beijing family and 8 strains belonging to non-Beijing family.Conclusions The Beijing family strains occupied a large proportion and the Beijing family MIRU genotype 223325173533 is a dominant strain in Linxia of Gansu.Combining detecting the deletion of RD105 and MIRU typing together provides a simple,fast,and effective method which is low in cost and might be practical and suitable for M.tuberculosis genotyping in China.