Urban air quality is influenced by vegetation through alterations in airflow and pollutant deposition processes. We investigated these interactions by integrating the Vegetation Impact Dynamic Assessment model (VIDA) with the Large-Eddy Simulation model PALM. Our analysis focus on nitrogen dioxide (NO₂) and particulate matter (PM) concentrations at the local scale, considering three tree genera. Our findings reveal the necessity of accounting for both gaseous pollutants and particles separately due to their differing mechanisms of deposition onto leaves. The coupled PALM-VIDA model demonstrates a significant reduction in PM levels across the modelling domain and within street canyons when deposition to vegetation is incorporated. Reduction in NO₂ through deposition to vegetation is lower but human NO2 exposure can still be decreased if tree species selection and placement leads to desirable effects on air flow. Sparse tree arrangements or species with sparse crowns facilitate ventilation and are often better at reducing NO₂ concentrations in street canyons compared to denser vegetation with higher deposition but negative effects on ventilation. Our study informs urban planning and green infrastructure design, underscoring the multifaceted role of urban greenery in air pollution mitigation strategies. Its main conclusion is that both deposition processes and the influence of air mixing and ventilation need to be considered to accurately assess the effects of urban trees on local air quality. Ill-considered placement and species selection may cause a net increase in pollutants underneath the trees. However, careful planning can address this risk and instead improve overall air quality.