This thesis presents a comprehensive analysis of the global health impacts of polycyclic aromatic hydrocarbons (PAHs) in ambient air, conducted on the basis of a high-resolution emission inventory, global chemical transport modeling, and probabilistic risk assessment. One of the main strengths of the thesis is the concentration downscaling process, which provides a linkage between emissions and exposure concentrations at a comparatively high resolution. Moreover, by focusing on individual susceptibility, the thesis proposes an instrumental revision of current risk assessment methodology and argues that, if individual susceptibility were not taken into consideration, the overall risk would be underestimated by 55% and the proportion of highly vulnerable populations would be underestimated by more than 90%.