A nested 3d-chemical transport model, GEOS-Chem, was used to estimate the burden of disease (BoD) from PM2.5 exposure in fifty-three urban agglomerations (UAs) in India with population of 1 million or more, and the contribution of different chemical components to the PM2.5 burden in these locations. Premature mortality due to four diseases (ischemic heart disease, IHD, chronic obstructive pulmonary disease, COPD, stroke and lung cancer) caused by exposure to PM2.5 was estimated using an integrated exposure response function. The studied 53 UAs accounted for approximately 20% of the total premature mortality of 0.54 (0.45–0.63) million in India. Delhi had the maximum premature mortality (11945) followed by Kolkata (9204), Mumbai (8817), Bangalore (4528) Hyderabad (4157) and Chennai (3818). Premature mortality was 2% to 59% higher than the country average in model regions containing UAs. Contribution from different chemical components was in the order: organic carbon (18–30%) > NO3- (8–21%) > SO42- (5–16%) > NH4+ (6–9%) > black carbon (2.5–4%). Our estimates suggest policies targeting multiple compounds are required to achieve large reductions in PM2.5 pollution in urban regions. © The Authors.