In this work, we study individual pollution exposure and its relationship with human mobility using large-scale behavioral data collected from smartphones and wristbands. In particular, we estimate PM2.5 exposure by employing GPS trajectories of approximately 500,000 individuals and global PM2.5 concentration data, aiming to explore the relationship between human mobility behavior and pollution exposure. We compare home-based and mobility-based exposure estimates and, following WHO 2021 guidelines, investigate how mobility-based risk changes with respect to residential estimates and conduct an individual stratification analysis to examine the role of socio-economic factors, aiming to identify high-risk classes of individuals and quantify disparities in pollution exposure. Finally, we analyze weekly patterns of mobility impact on exposure to examine how exposure differs between weekdays and weekend.