This paper systematically analyzes the characteristics of spatial and temporal distribution of NO2 vertical column density from 2005 to 2017 in Chengdu-Chongqing Urban Agglomeration (CCUA). The tropospheric NO2 vertical column density data were from the latest version of OMI (Ozone Monitoring Instrument) sensor of Aura satellite. The techniques of linear fitting, trend analysis and spatial autocorrelation analysis were used in this paper. The results showed that in CCUA there was reasonable correlation between the ground monitoring NO2 column density and the data from OMI sensor, and their variation trends were basically consistent. From 2005 to 2012, the annual NO2 column density increased about 52.38%, and it decreased after 2012, and for the density level in four seasons, winter > autumn > spring > summer. The monthly averaged values showed obvious periodicity, which was expressed in terms of sinusoidal function. The cycle was about one year, with the lowest value in July or August, and the highest value in November, December or January;Spatially, the NO2 column density in thecomparatively developed areas were higher than in the economic less-developed areas. From 2005 to 2012, for about 99.58% of the total area, the NO2 column density increased and from 2012 to 2017. for about 80.50% of the total area, the NO2 column density decreased. In general, from 2005 to 2017, in the eastern and western mountainous areas of CCUA, the NO2 column density showed an increasing trend, and there was a significantly positive spatial correlation in Chengdu and Chongqing.