BACKGROUND: Interferon-gamma (IFN-gamma) production by lymphocytes exposed to antigens specific of Mycobacterium tuberculosis has been shown to correlate with antigen load and disease activity. AIM OF STUDY: To determine whether treatment of tuberculosis (TB) led to a decrease and/or a reversion of results of a IFN-gamma release assay (T-SPOT.TB, Oxford Immunotec, UK) and thus if T-SPOT.TB could be used to monitor response to treatment. METHODS: Qualitative and quantitative analysis (SFUs: spot-forming units) of T-SPOT.TB in HIV-negative patients with TB, during initial 2 weeks of treatment (T0), at end of treatment (TE) and 6 months later (TE+6). RESULTS: Mean SFU (SD) was 75 (58; n=62) at T0, 46 (55; n=55) at TE, and 33 (46; n=41) at TE+6; positive rate was 98%, 93% and 98%, respectively. SFUs (paired samples, n=36) decreased significantly between T0 and TE; 2 reversions occurred between T0 and TE (6%), but none between TE and TE+6. Of 6 patients (17%) with an increase in SFUs between T0 and TE, 5 had a favourable outcome at TE and TE+6. CONCLUSION: Decrease in SFUs under treatment suggests a relationship with antigen load; however, persisting high SFUs were not predictive of unfavourable outcome and test reversion was rare.