Vegetation barriers are commonly regarded as an effective strategy in urban planning for mitigating traffic-induced airborne pollutants and associated exposure of road-users and urban dwellers. Locally verified evidence is needed to better understand the nexus between vegetation barrier designs and PM_2.5 dispersion in open-road environments and provide practical insights for the creation of a healthy urban environment. In this study, a 3D numerical model is constructed and validated using in-situ data collected via field campaign to simulate PM_2.5 dispersion in a sidewalk flanked by building structure with varying planting designs of vegetation barriers (short vs. tall bushes in different location and coverage). The simulation results reveal that vegetation barrier could contribute to PM_2.5 mitigation, with an average reduction of PM_2.5 contribution ranging between 15.95 µg/m³ and 20.38 µg/m³ on the leeward side of the vegetation barrier. Bushes, regardless their size, tend to be more effective than arbors. The building structure flanked the sidewalk can also influence PM_2.5 concentrations in the sidewalk canyon. Various design scenarios using short bushes could consistently reduce PM_2.5 concentrations on both sides of the sidewalk. For design scenarios using tall bushes, PM_2.5 concentration reduction at the leeward side of the vegetation barrier can be further enhanced by the presence of building structure, while PM_2.5 concentrations along the vertical dimension of building facade could be reduced only when 1/3 bush design scenarios are adopted, which could be attributed to the formation of vortex circulations in the sidewalk canyon. Therefore, constructing vegetation barriers with a large proportion of short bushes may be a practical solution to minimizing the exposure of urban dwellers to traffic-induced pollutants in open-road environments.