Interferon-gamma release assays (IGRAs) provide a new diagnostic method for Mycobacterium tuberculosis (TB) infection. However, the diagnostic value of IGRAs for extrapulmonary TB (EP-TB) has not been clarified. We evaluated its diagnostic usefulness in patients suspected with EP-TB. Of the 102 patients with suspected EP-TB, 48 were diagnosed active EP-TB. Of the 48 EP-TB patients, 16 (33.3%) were diagnosed with TB pleuritis, 14 (29.2%) with TB lymphadenitis, 10 (20.8%) with skeletal TB and 7 (14.6%) with TB enteritis. The overall sensitivity and specificity of the assay were 75.0% (95% confidence interval (CI), 64.2-83.7%) and 75.9% (95% CI, 66.3-83.6%), respectively. Among 40 patients presented with pleuritis, the QuantiFERON-TB Gold In-Tube (QFT-IT) assay showed 43.8% (95% CI, 23.8-61.0%) sensitivity and 79.2% (95% CI, 65.9-90.7%) specificity, whereas in 15 patients with skeletal involvement, the QFT-IT assay showed 90.0% (95% CI, 68.0-99.1%) sensitivity and 80.0% (95% CI, 36.0-98.2%) specificity, respectively. Also in 22 patients with cervical lymphadenopathy, the QFT-IT assay showed 92.9% sensitivity and 75.0% specificity, and in 12 patients with GI involvements, it showed 85.7% sensi-tivity and 80.0% specificity. These findings suggest that diagnostic performance of the QFT-IT assay may vary depending on the site of TB involvement. We suggest that QFT-IT is helpful for the diagnosis of TB lymphadenopathy, skeletal TB and TB enteritis, but it is not sufficient for the diagnosis for TB pleuritis.