Epidemiological studies show heterogeneities in the particulate pollution-related exposure-effect relationships among cardiorespiratory patients, but the connection to chemical composition and toxic properties of the inhaled particles is largely unknown. To identify the chemical constituents and sources responsible for the diverse inflammatory and cytotoxic effects of urban air, fine (PM(2.5-0.2)) and coarse (PM(10-2.5)) particulate samples were collected during contrasting air pollution situations. We exposed mouse RAW 246.7 macrophages for 24 hrs to PM(2.5-0.2) and PM(10-2.5) samples from six European cities. The concentrations of proinflammatory cytokines (IL-6, TNFalpha), chemokine (MIP-2), and nitric oxide were measured from the cell culture medium, and the cytotoxicity was assayed. Spearman's correlations between the chemical constituents and cellular responses were analyzed. In the PM(2.5-0.2) size range, the tracers of photo-oxidation of organics in the atmosphere (oxalate, succinate, malonate), some transition metals (Ni, V, Fe, Cu, Cr), and insoluble soil constituents (Ca, Al, Fe, Si) correlated positively with the response parameters. In contrast, the tracers of incomplete biomass (monosaccharide anhydrides) and coal (As) combustion, and polycyclic aromatic hydrocarbons (PAHs), had negative correlations with the inflammatory activity. The compositions of PM(10-2.5) samples were more uniform and there were only occasional high correlations between the chemical constituents, endotoxin, and the response parameters. The present results suggest that the local sources of incomplete combustion and resuspended road dust are important producers of harmful fine particulate constituents that may, however, operate via diverse toxicity mechanisms. The results agree well with our recent findings in the mouse lung.