Background & Objective. IRIS is an important complication that occurs during management of HIV-TB coinfection and it poses difficulty in diagnosis. Previous studies have reported variable incidence of IRIS. The present study was undertaken to describe the pattern of TB-associated IRIS using recently proposed consensus case-definitions for TB-IRIS for its use in resource-limited settings. Methods. A prospective analysis of ART-naïve adults started on HAART from November, 2008 to May, 2010 was done in a tertiary care hospital in north India. A total 224 patients divided into two groups, one with HIV-TB and the other with HIV alone, were followedup for a minimum period of 3 months. The diagnosis of TB was categorised as ‘‘definitive’’ and ‘‘probable’’. Results. Out of a total of 224 patients, 203 completed followup. Paradoxical TB-IRIS occurred in 5 of 123 (4%) HIV-TB patients while 6 of 80 (7.5%) HIV patients developed ART-associated TB. A reduction in plasma viral load was significantly ( 𝑃 = . 0 1 6 ) associated with paradoxical TB-IRIS. No identifiable risk factors were associated with the development of ART-associated TB. Conclusion. The consensus case-definitions are useful tools in the diagnosis of TB-associated IRIS. High index of clinical suspicion is required for an early diagnosis.