To characterize the elemental composition and source apportionment of aerosols in roadside area, particulate matters with aerodynamic diameter less than 2.5 μm (PM 2.5 ) were collected in close proximity to a road from September 2017 to February 2018 in downtown Chengdu, China. An energy-dispersive X-ray fluorescence spectrometer was used to quantify elemental constituents (Al, Si, S, K, Mn, Fe, Ni, Cu, Zn, As, and Pb) of PM 2.5 and was calibrated by inhouse standards instead of commercial standards. The constructed calibration curves exhibited good linearity with all correlation coefficients greater than 0.98. The proposed calibration method proved to be reliable for the subsequent elemental analyses due to the satisfactory performance of u-score and precision that were validated by the certified reference materials (#2783). The results revealed that average PM 2.5 concentrations of 92.2 ± 45.6 and 113.2 ± 60.3 μg/m 3 were respectively observed in autumn and in winter. The major trace elements identified were K, S, and Fe and the minor contributions were from Cu and As. Most crustal elements showed decline in winter except for K, and most anthropogenic elements showed increase in winter except for Ni. Using rotation factor analysis and cluster analysis based on the elemental dataset, four potential sources were identified: road dust, vehicular emissions, coal and biomass burning, and industrial emissions. This research will provide a better understanding of traffic-related PM 2.5 composition, and this can be used in the mitigation and prevention programs.