Roof greening is a new technique for improvement of outdoor thermal environment which influences air quality through its impacts on thermal and flow field. In order to examine effects of green roofs on reactive pollutant dispersion within urban street canyons, a computational fluid dynamics (CFD) model was employed which contained NO-NO2-O3 photochemistry and energy balance models. Simulations were performed for street canyons with different aspect ratios (H/W) of 0.5, 1.0, and 2.0 such that leaf area density (LAD) of green roofs changed. It was found that roof greening led distribution of pollutants to alter forH/W = 0.5 and 1.0 cases in such a manner that their averaged concentrations had small variations as LAD changed. However, by increasing LAD inH/W = 2.0, ventilation efficiency of nitrogen oxides increased since the flow was enhanced within the canyon. Additionally, averaged concentration of ozone inH/W = 2.0 increased with increasing LAD, owing to downward flux of ozone at roof level. Results show that roof greening is a good strategy which can be used in order to improve air quality and thermal environment, especialy within deep street canyons.