Pedestrian overpasses improve urban mobility but can create localized hotspots of traffic-related particulate matter (TRPM) exposure. This study combines multi-season field measurements and process-based dispersion modelling to quantify how overpass design and surrounding street morphology affect PM_1.0, PM_2.5, and black carbon (BC) distributions on pedestrian overpasses. Two representative Ⅰ-shaped overpasses in Fuzhou, China were monitored across three seasons using 10-s mobile measurements at pedestrian breathing height, covering both upper-level walkways and traffic-adjacent entrances in open-type and canyon-type settings. Field measurements consistently identified elevated concentrations along upper corridors and near entrances. Seasonally, the open-type overpass showed 17.1% higher PM_1.0 in autumn, while the canyon-type overpass exhibited 9.3% higher BC in winter, indicating contrasting dispersion constraints. Generalized additive models diagnosed conditions linked to high TRPM, highlighting low wind speeds and unfavorable wind directions as key drivers. Guided by these regimes, coupled traffic–emission–dispersion simulations evaluated design sensitivity. Scenario results indicate that widening streets relative to building height and increasing deck elevation from 5 to 7 m reduce PM_2.5 by approximately 15%–17% and BC by approximately 10%–45%. Partial side shielding lowered PM_2.5 by 31.8% and modestly reduced BC by 4.5%, whereas full enclosure increased PM_2.5 by 41.2% and BC by 91.9% due to restricted ventilation. These findings highlight street–infrastructure interactions as a practical lever for reducing pedestrian TRPM exposure along major urban arterials.