Objective The aim of this study was to evaluate the diagnostic performance of T-SPOT.TB for tuberculous lymphadenitis. Methods Suspected tuberculous lymphadenitis patients between September 2010 and September 2018 who had both peripheral blood T-SPOT.TB test and lymph node biopsy were retrospectively enrolled in this study. The cutoff value of T-SPOT.TB test for peripheral blood was set as 24 spot forming cell (SFC)/10 6 periphreral blood monocyte cell (PBMC) according to the instruction of testing kits. The gold standard for diagnosis of TBL was the combination of microbiology results, histopathology results and patient's response to anti-TB treatment. Diagnostic efficacy of T-SPOT.TB was evaluated, including sensitivity, specificity, accuracy, predictive values, and likelihood ratio. Results Among 91 patients who met the inclusion criteria, we excluded 8 cases with incomplete clinical information and 6 cases who lost to follow-up. According to the gold standard, there were 37 cases of true TBL (9 confirmed TBL and 28 probable TBL), 30 cases of non-TBL, and 10 cases of clinically indeterminate diagnosis who were excluded from the final analyses. The T-SPOT.TB tests yielded 43 cases of positive response and 24 cases of negative response. The sensitivity, specificity, accuracy, positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (PLR) and negative likelihood ratio (NLR) of peripheral blood T-SPOT.TB for diagnosing TBL were 89.2%, 66.7%, 79.1%, 76.7%, 83.3%, 2.68 and 0.16, respectively. The number of SFCs of T-SPOT.TB in TBL patients [432(134-1264)/10 6 PBMCs] was higher than that in non-TBL patients [0 (0-30) /10 6PBMCs] with a significant difference (Z=-5.306, P <0.001).Conclusion T-SPOT.TB is a rapid and simple diagnostic test for TBL with a high sensitivity and negative predictive value.