This study explored the capabilities offered by the aerosol optical depth (AOD) product (created on a daily basis from MODIS (Moderate Resolution Imaging Spectroradiometer) images acquired at 10:30 AM and 1:30 PM local time, to predict the at-surface concentrations of fine particulate matter (PM2.5) measured on the monitoring stations of the city of Santiago, Chile, during free cloud days of year 2016. To accomplish that, linear correlations were performed between remote-based and field-based data sources, both matched in time and space, in order to find the days and stations where MODIS-AOD may offer the highest correlation. Based on these cases, a linear regression model was performed in order to explore the predictive power of MODIS-AOD. The highest correlations were consistently obtained when MODIS-AOD data created at 1:30 PM were used at the station level for all the sampled days of the study period. When the monitoring stations of Cerro Navia, Quilicura and Pudahuel were grouped, the correlation coefficient was 0.77 and the predictive model explained the 63% of the at-surface MP2.5 concentrations and yielded an accuracy of 69.1%. The findings of this exploratory study provide a first approach to the modeling of the at-surface concentrations of fine particulate matter in the city of Santiago, which could be improved in a future with the aim to anticipate environmental warnings and disease outbreaks related to the exposure of this pollutant.