Background: Lymph node tuberculosis (LNTB) is the leading type of extrapulmonary tuberculosis (EPTB) causing death in children. The Xpert MTB/RIF assay is a novel rapid test for the diagnosis of LNTB. Although previous evidence suggests that Xpert is reliably accurate in diagnosing EPTB in children, information is lacking for the specific type of LNTB in children. The aim of this study was to systematically assess the accuracy and reliability of Xpert for the diagnosis of LNTB in children. Methods: We systematically searched four databases, Embase, Cochrane Library, PubMed, and Web of Science, which extracted relevant data according to predefined inclusion and exclusion criteria. The data were analyzed by meta-Disc 1.4 and Stata 12.0 software to determine sensitivity, specificity, diagnostic odds ratio (DOR), etc. Results: A total of 646 samples from 8 studies were included in the analysis. The pooled sensitivity, specificity, negative likelihood ratio (NLR), positive likelihood ratio (PLR,) and combined diagnostic odds ratio (DOR) of Xpert for all samples were 0.79 (95% CI 0.70, 0.87), 0.90 (95% CI 0.86, 0.92), 0.29 (95% CI 0.19, 0.43), 7.20 (95% CI 3.32, 15.60), and 37.56 (95% CI 13.04, 108.15), respectively. The area under the curve (AUC) of the summary receiver operating characteristic (sROC) curve was 0.9050. Conclusion: Overall, Xpert showed moderate sensitivity and high specificity compared with culture in the diagnosis of LNTB in children. In addition, after analyzing the combined diagnostic odds ratio and positive LR, our study showed that Xpert has excellent diagnostic accuracy.