Abstract: The authors selected 35 large and medium-sized cities in China as sample cities to evaluate the air quality and composition of urban development land characteristics in 2016. The relationship between composition of urban development land and air quality was analyzed by using canonical correlation analysis method, which provides a scientific basis for rational arrangement of urban land structure, improvement of air quality and urban human settlements quality. The results showed that: (1) the regional differences of air quality in sample cities varied significantly during the study period. The air pollution was mainly concentrated in north and northwest of China, while the air quality was relatively good in southeast coast. Compared with other cities in China, air pollution in Beijing-Tianjin-Hebei area was still very serious. The average annual concentration of air quality index, PM2.5, PM10, SO2, NO2 showed that the southern city were better than the northern city and the coastal city were better than the inland city. (2) During the study period, there were some differences in the composition of urban development land, and residential land and industrial and mining storage land account for 49%- 58% of the total area. The structural differences between northern and southern cities was mainly reflected in the proportion of green space and square land and the differences of composition of urban development land in coastal, inland cities was mainly reflected in the proportion of public management and public service land, industrial and mining storage land and green space and square land. The average value of information entropy showed that the northern cities were greater than southern cities, inland cities were greater than coastal cities. (3) The results of canonical correlation analysis showed that the composition of urban development land had a certain influence on air quality. Air pollutants in different regions were different and the structural factors that affected the concentration of air pollutants were different. On the whole, there was a negative correlation between residential land, commercial service facility land and SO2; a positive correlation between industrial and mining storage land and NO2. Regionally, residential land, road and transportation facilities land were positively correlated with PM2.5; industrial and mining storage land was positively correlated with NO2; residential land, transportation facilities land were negatively correlated with SO2. Composition of urban development land of sample cities has a certain impact on air quality.