Rheumatoid arthritis (RA) is a systemic autoimmune disease in which inflammation of the joints is one of the dominant clinical abnormalities resulting in serious morbidity. Over the past decade, tumour necrosis factor (TNF) antagonist has revolutionized the treatment of RA. However, the subsequent increased risk of developing tuberculosis is one of the major drawbacks of this otherwise effective treatment. Latent tuberculosis infection (LTBI) is an asymptomatic form of tuberculosis that is confined by the host's immune system. Active tuberculosis may develop when the immune status weakens. This risk is much higher in patients receiving TNF antagonist. Traditionally, tuberculin skin test (TST) is used to diagnose LTBI. Unfortunately, TST cannot distinguish bacillus Calmette-Guérin (BCG) vaccination from tuberculosis making it difficult to use as a reliable diagnostic tool. In addition, possible anergy and interaction of the altered autoimmune status in rheumatological diseases further complicate the interpretation of TST results. Although interferon-gamma release assay (IGRA) has improved the diagnosis of LTBI in immunocompetent individuals, its respective sensitivity/specificity values are unknown in patients with autoimmune disease due to variable pretest probability and lack of confirmatory test for LTBI. Thus, the use of IGRA for screening LTBI is variable among different countries. This review explores the prevalence of tuberculosis in patients receiving TNF antagonist in countries with different tuberculosis disease burdens and the potential mechanisms for variation in the incidence of tuberculosis with different TNF antagonists, the current practice guidelines for assessing the risk of LTBI in different countries, and the possible solutions for improving diagnosis, monitoring and management of LTBI.