The aim of the study was to compare the usefulness of the QuantiFERON-TB Gold (QFT-2G) with that of the tuberculin skin test (TST) for detecting previous infection of tuberculosis (TB) in Japanese rheumatoid arthritis (RA) patients. Before receiving biologic therapy, 97 RA patients were divided into two groups based on their chest computed tomography (CT) findings: the TB past infection group (n = 48), with old inflammatory changes due to prior pulmonary TB; and the non-TB infection group (n = 49), without such findings. The QFT-2G was not affected by methotrexate or prednisolone. Indeterminate results with a positive control had a low incidence (5.2%). A positive QFT-2G for the TB past infection group at cutoffs of 0.35 and 0.1 IU/ml (intermediate range) was seen in 5.8% and 20.8%, respectively. A TST >20 mm was significantly higher in the non-TB infection group (31%) than in the TB past infection group (13%). The correlation between the QFT-2G and TST was poor among all patients. Disagreement between these tests in the non-TB infection group was caused by the false-positive TST induced by previous Bacillus Calmette-Guérin (BCG) vaccination. Only 12 (12.4%) of 97 patients had a positive QFT-2G (≥0.1 IU/ml) and a negative TST (<20 mm), but in this subgroup, a high incidence (10, 83.3%) was detected in the TB past infection group. QFT-2G may be a good alternative to the TST to evaluate previous TB infection when it is necessary to determine whether isoniazid (INH) prophylaxis is needed before biologic therapy is begun.