This study numerically investigates the influence of different vegetation types and layouts on microclimate and air quality in residential districts based on the morphology and green layout of Nanjing, China. Simulations were performed using Computational Fluid Dynamics and the microclimate model ENVI-met. Four green indices, i.e., the green cover ratio, the grass and shrub cover ratio, the ecological landscaping plot ratio and the landscaping isolation index, were combined to evaluate thermal and wind fields, as well as air quality in district models. Results show that under the same green cover ratio (i.e., the same quantity of all types of vegetation), the reduction of grass and shrub cover ratio (i.e., the quantity of grass and shrubs), replaced by trees, has an impact, even though small, on thermal comfort, wind speed and air pollution, and increases the leisure space for occupants. When trees are present, a low ecological landscaping plot ratio (which expresses the weight of carbon dioxide absorption and is larger in the presence of trees) is preferable due to a lower blocking effect on wind and pollutant dispersion. In conjunction with a low landscaping plot ratio, a high landscaping isolation index (which means a distributed structure of vegetation) enhances the capability of local cooling and the general thermal comfort, decreasing the average temperature up to about 0.5 °C and the average predicted mean vote (PMV) up to about 20% compared with the non-green scenario. This paper shows that the relationship vegetation-microclimate-air quality should be analyzed taking into account not only the total area covered by vegetation but also its layout and degree of aggregation.