The prevalence of tuberculosis (TB) in the elderly is higher than that in the general population, and elderly populations are considered a high-risk group. Currently, annual TB screening of Bacille Calmette-Guérin (BCG)-vaccinated people aged over 65 years is performed by an annual chest x-ray examination (CXR) in Japan. Interferon-gamma release assays (QuantiFERON-TB Gold and QuantiFERON-TB Gold In-Tube [QFT]) are new alternatives to the tuberculin skin test to diagnose latent TB infection (LTBI) that have no cross-reactivity with the BCG vaccine. We evaluated the cost effectiveness of QFT versus CXR versus no screening in BCG-vaccinated elderly populations. We constructed a Markov model to evaluate the cost effectiveness of QFT, CXR, and no screening. The target population was a hypothetical cohort of 1000 immunocompetent 65-year-olds, using a societal perspective and a lifetime horizon. All costs and clinical benefits were discounted at a fixed annual rate of 3%. In the base-case analysis, a no-screening strategy resulted in the lowest cost ($US303.51; 14.6475 quality-adjusted life-years [QALYs]) compared with CXR ($US393.22; 14.6477 QALYs) and QFT ($US525.45; 14.6516 QALYs) [year 2008 values]. The sensitivity of QFT, as well as the prevalence of TB and LTBI, influenced the cost effectiveness; when the sensitivity of QFT was higher than 0.89, QFT became more cost effective than providing no screening. As the prevalence of LTBI and TB increased, the QFT strategy became progressively more cost effective. Providing no routine TB screening is currently the most cost-effective strategy for BCG-vaccinated elderly populations in Japan. There appears to be little role for CXR in TB screening of elderly populations. These findings may be applicable to other countries with intermediate and high TB risks when choosing optimal TB screening of elderly populations.