We assessed mammalian cell cytotoxicity of ambient PM2.5 and investigated its association with the oxidative potential (OP) and chemical composition of the particles. Sixteen PM samples spanning in various seasons (fall, winter, spring and summer) were collected from an urban site in central Illinois. Cytotoxicity (LC50) in terms of the volume of air that kills 50% of the cells were calculated, which varied from 4.3 to 7.2 m3 of air. The OP was measured by two assays – the dithiothreitol (DTT) and the surrogate lung fluid (SLF) assay. In DTT assay, we measured two endpoints – hydroxyl radicals (•OH) generation and DTT consumption (the conventionally measured endpoint), while only •OH generation was measured in the SLF assay. Although, all three endpoints in the OP assays correlated significantly (P ≤ 0.05) with LC50, the correlation of reactive oxygen species (ROS) generation in DTT and SLF assays was much higher (r > 0.80 for •OH generation versus LC50) than the DTT consumption (r = 0.58). To further understand the components in PM that drive cytotoxicity and OP, concentration of water-soluble metals (Fe, Cu, Co, Cr, Mn, Ni, Pb, V, Hg, and Zn), organic carbon (OC), water soluble organic carbon (WSOC), and elemental carbon (EC) were measured. Among all the chemical components, Fe, Cu and WSOC correlated most (r > 0.70; P ≤ 0.01) with the cytotoxicity. DTT consumption correlated only with OC and WSOC (r > 0.80; P ≤ 0.01), while •OH generation in DTT and SLF assay correlated with both WSOC (r > 0.70; P ≤ 0.01) and metals (i.e. Fe and Cu; r > 0.75; P ≤ 0.01). Our results suggest a strong link between the PM2.5 OP and its cytotoxicity. Furthermore, the synergistic interactions among the organic compounds (i.e. WSOC) and metals (Fe and Cu) to enhance the ROS generation, which are more effectively captured in •OH generation endpoints in DTT and SLF assay than the DTT consumption, appear to be largely responsible for the observed mammalian cell cytotoxicity of PM2.5.