Contributions to PM2.5 from different emission sectors over central Chile and the Santiago metropolitan area during summer/fall and winter have been studied using a chemistry transport model. Simulated mean concentrations of PM2.5 are generally underestimated when compared to measurement stations in Santiago forming part of the Chilean National Air Quality Information System (SINCA). Reasons for this could include underestimated direct PM2.5 emissions; missing emissions of semi- and intermediately-volatile organic compounds, SVOC and IVOC; and too high wind speeds in the simulations. Simulated PM2.5 concentrations in Santiago in winter are lower than observed during night-time and higher during daytime and late evening which can be related to too high simulated wind speeds, as well as to uncertainties in the diurnal variation of the emissions. In summer the diurnal variation is closer to the observed, but the peak concentrations in the mornings are underestimated, and concentrations in the evening are overestimated. Simulated contributions from different aerosol components to PM2.5 at one station in Santiago are lower than observed for all components except for elemental carbon equivalent black carbon, BCe, for which simulated concentrations are comparable or higher than observed. The absolute differences are largest for total organic matter and relative differences are largest for BCe and ammonia. Simulated sector contributions show that emissions from transport and construction machinery dominate in Santiago while residential wood combustion dominates in other urban areas in central Chile except close to major point sources. Secondary inorganic aerosol is estimated to be the largest aerosol component away from urban areas, traffic routes and major industrial sources, while modelled secondary organic aerosol gives a small contribution.