The aim of this study was to identify the sources of atmospheric pollutants in densely populated urban areas from a particle toxicity perspective. To this end, the Positive Matrix Factorization (PMF) model and vehicle flux analysis were used to identify the sources of atmospheric pollutants in an urban area based on the measured compounds and wind speed at the receptor site. Moreover, the toxicity of each emission source was compared with the dithiothreitol-oxidation potential normalized to 9,10-Phenanthrenequinone (QDTT-OP) analysis using the PMF source apportionment results. The study found that the dominant sources of atmospheric pollutants in the urban area examined were secondary product (43.7%), resuspended dust (25.4%), and vehicle emissions (14.4%). The vehicle flux analysis demonstrated that reducing the number of vehicles could directly reduce urban atmospheric pollutants. By comparing the time series of PMF source profiles with QDTT-OP, the QDTT-OP analysis showed an r2 value of 0.9, thus indicating a strong correlation with biomass burning as the most harmful source of PM2.5 based on emission sources. Overall, this study is expected to provide valuable guidance for managing atmospheric pollutants in densely populated urban areas, and the findings could serve as a helpful resource for improving urban air quality in the future.