Santiago (33.5°S, 70.5°W), the capital of Chile, is frequently affected by extreme air pollution events during wintertime deteriorating air quality (AQ) and thus affecting the health of its population. Intense residential heating and on-road transport emissions combined with poor circulation and vertical mixing are the main factors responsible for these events. A modelling system composed of a chemistry-transport model (CHIMERE) and a meteorological model (WRF) was implemented to assess the AQ impacts of residential and transportation sources in the Santiago basin. A two-week period of July 2015 with various days with poor AQ was simulated focusing on the impact on AQ with respect to fully inhalable particles (PM2.5) and nitrogen oxides (NOX). Three emission scenarios, within the range of targeted reductions of the decontamination plan of Santiago, were tested; namely 50% reduction of residential emission, 50% reduction of transport emissions and the combination of both. An additional scenario decreasing transport emissions in 10% was carried out to examine whether a linear dependence of surface concentrations on changes in emissions exists.The system was validated against surface and vertically resolved meteorological measurements. The model reproduces the daily surface concentration variability from the AQ monitoring network of Santiago. However, the model not fully captures the emissions variations inferred from the observations which may be due to missing sources such as resuspension of dust. Results show that, during the period studied, although both residential and transportation sources contribute to observed AQ levels in Santiago, reducing transport emissions is more effective in terms of reducing the number of days with pollution events than decreasing residential combustion. This difference in impact is largely due to the spatial distribution of the emission sources. While most of the residential combustion is emitted in the outskirts of the city, most of the transport emissions occur within the city, where most of the stations from AQ monitoring network of Santiago are located. As can be expected, the largest improvement of AQ in Santiago is achieved by the combined reduction of emissions in both sectors. Sensitivity analysis with 10% reduction in transport emissions reveals a linear behavior between emissions and concentrations for NOX and approximate linear behavior for PM2.5. The absence of secondary aerosols formation and dust resuspension in the current simulation could explain this deviation from linearity for fine particles. Nevertheless, it suggests that the results can be used for mitigation policies with emissions reductions below the 50% used in this study.