throughout the world. Particulate matter (PM) is a criteria pollutant that is of high interest in urban locations. The precise characteristics of PM in a given locale depend on the source origin, which in turn is a function of economic, social and technological factors. In order to effectively manage PM and thereby, the exposure risk to humans, it is very essential to identify the main sources and their contributions from source emissions. Receptor modelling plays a major role in identifying and apportioning sources of airborne PM across the world. Unmix model is a multivariate receptor model developed by the United States Environmental Protection Agency (U.S.EPA) based on factor analysis, which estimates the number of sources using a singular value decomposition method to reduce the dimensionality of data. In this study, Unmix receptor model version 6.0 is used to identify and quantify the sources of PM at Chennai; a metropolis in southern India. A total of 29 elements (Ag, Al, As, B, Ba, Be, Bi, Ca, Cd, Co, Cr, Cu, Fe, Ga, K, Li, Mg, Mn, Mo, Na, Ni, Pb, Rb, Se, Sr, Te, Tl, V and Zn) and ten ions (Na+, NH4 +, K+, Ca2+, Mg2+, F-, Cl-, NO2-, NO3 and SO4+2-) were analysed to find the chemical characteristics of PM10 and PM2.5. Four sources were identified for both PM10 and PM2.5. Vehicular pollution (11%), crustal source (27%), marine aerosol (40%) and industrial source (22%) are the sources identified for PM10. Vehicular emissions (32%), secondary aerosol (13%), marine aerosol (33%) and industrial source (22%) are the sources identified for PM2.5. © 2019 Technoscience Publications. All rights reserved.