Background: Tuberculosis (TB) is common among HIV-infected individuals in many resource-limited countries and has been associated with poor survival. We evaluated morbidity and mortality among individuals first starting antiretroviral therapy (ART) with concurrent active TB or other AIDS-defining disease using data from the "Prospective Evaluation of Antiretrovirals in Resource-Limited Settings" (PEARLS) study. Methods: PARTICIPANTS WERE CATEGORIZED RETROSPECTIVELY INTO THREE GROUPS ACCORDING TO PRESENCE OF ACTIVE CONFIRMED OR PRESUMPTIVE DISEASE AT ART INITIATION: those with pulmonary and/or extrapulmonary TB ("TB" group), those with other non-TB AIDS-defining disease ("other disease"), or those without concurrent TB or other AIDS-defining disease ("no disease"). Primary outcome was time to the first of virologic failure, HIV disease progression or death. Since the groups differed in characteristics, proportional hazard models were used to compare the hazard of the primary outcome among study groups, adjusting for age, sex, country, screening CD4 count, baseline viral load and ART regimen. Results: 31 of 102 participants (30%) in the "TB" group, 11 of 56 (20%) in the "other disease" group, and 287 of 1413 (20%) in the "no disease" group experienced a primary outcome event (p = 0.042). This difference reflected higher mortality in the TB group: 15 (15%), 0 (0%) and 41 (3%) participants died, respectively (p<0.001). The adjusted hazard ratio comparing the "TB" and "no disease" groups was 1.39 (95% confidence interval: 0.93-2.10; p = 0.11) for the primary outcome and 3.41 (1.72-6.75; p<0.001) for death. Conclusions: Active TB at ART initiation was associated with increased risk of mortality in HIV-1 infected patients.