Objective To investigate the diagnostic value and influencing factors of Xpert MTB/RIF assay （Xpert） for active tuberculosis in children. Methods The sensitivity and specificity were compared among tuberculin skin test （TST）， interferon gamma release assays （IGRAs） and Xpert among different ages and types of active tuberculosis in children， and the sensitivity of Xpert was compared among different types of specimens. Results A group of 231 children with active tuberculosis and another group of 216 without tuberculosis were included in the study. The sensitivity of Xpert for active tuberculosis was 62.78%， lower than that of IGRAs （P < 0.05）， and the specificity was 100%， higher than those of IGRAs and TST （both P < 0.001）. The sensitivity of Xpert was lower than that of IGRAs in the children younger than 5 years old （74.71% vs. 67.47%， P > 0.05）， while the sensitivity of Xpert was significantly lower than that of IGRAs in the children older than 5 years old （58.26% vs. 84.62%， P < 0.001）. The sensitivity of the pulmonary tuberculosis group was significantly higher than that of the extrapulmonary tuberculosis group （P < 0.001）. There were statistical differences among the different types of specimens （P < 0.05），With the highest sensitivity of Xpert detected with bronchoalveolar lavage fluid （BALF）. For the suspected pulmonary tuberculosis， the sensitivity of Xpert tested in the BALF was higher than that tested in the sputum （P = 0.012）. For the suspected pulmonary plus extrapulmonary tuberculosis， the sensitivity of Xpert tested in therespiratory specimens was higher than that tested in the non-respiratory specimens （P = 0.004）. Conclusion Xpert is effective in diagnosing childhood active tuberculosis， especially used in testing pulmonary tuberculosis and BALF of those suspected children younger than 5 years old. For pulmonary plus extrapulmonary tuberculosis， Xpert with respiratory tract specimens are first recommended.