Street-level concentrations of nitrogen dioxide (NO₂) and particulate matter less than 10 μm in aerodynamic diameter (PM₁₀), exceed public health standards in many cities, causing increased mortality and morbidity. Concentrations can be reduced by controlling emissions, increasing dispersion, or increasing deposition rates, but relatively little attention has been paid to the latter as a pollution control method. Both NO₂ and PM₁₀ are dry deposited onto surfaces at rates that vary according to the nature of the surface; typically low for urban surfaces such as brick, but much higher for vegetated surfaces. Previously, city-scale studies have suggested that deposition to urban vegetation makes very modest improvements to urban air quality; less than 5% even when the surface is fully covered by trees. However, these studies do not take account of the interplay between urban form and vegetation. Street canyons limit mixing between street-level air and the overlying urban boundary layer (UBL), effectively trapping air within them, an effect which increases as the aspect (height/width) ratio increases. As a result, increasing deposition within street canyons is far more effective at reducing pollutant concentrations than increasing deposition directly from the UBL. By developing and applying a new model of deposition, mixing and chemistry within street canyons, we show that increasing deposition by the planting of vegetation in street canyons, particularly green walls, can reduce street-level concentrations by as much as 40% for NO₂ and 60% for PM₁₀. Our results suggest that the further limitations in mixing provided by street trees (or increases in canyon aspect ratio) in many cases increase the deposition efficiency sufficiently that they yield an improvement in in-canyon air quality, counter to the results of mixing-only studies. Hence, we believe the benefits of vegetation for air quality have been substantially undervalued. Importantly, whereas previous studies have relied on large-scale urban greening to achieve a significant air quality improvement, we show that these substantial local improvements can be gained through action at the scale of a single street canyon alone. Our results show that judicious use of vegetation promises to create an efficient urban pollutant filter, yielding rapid and sustained improvements in urban air quality, and highlight the importance of further research into the interactions of vegetation with the urban atmosphere. lity, and highlight the importance of further research into the interactions of vegetation with the urban atmosphere.