Abstract: Urban area is a significant contributor to the global carbon dioxide emissions. In order to understand the influences of meteorological background conditions, land use and land cover, as well as energy consumption on urban CO_2emission, temporal and spatial distribution characteristics and annual emissions were analyzed, based on observed data obtained by eddy-covariance measurements at 140m above the ground on Beijing Meteorological Tower in a high-density residential area from June 26^th, 2009to December 31^st, 2011.The results showed that the diurnal patterns of CO_2were largely influenced by traffic volume and typical two-peak traffic patterns. Moreover, the influence was particularly prominent in winter, since fuel consumption for heating during winter considerably enhances CO_2emission. Due to heating and plant dormancy, the daily average flux measured in winter (30μmol/(m²·s)) was significantly higher than those in other seasons with 15.2,7.9and 15.8μmol/(m²·s) in spring, summer and autumn, respectively (t-test, P<0.001). Spatial variation patterns of CO_2flux were mainly determined by the prevailing surface cover within the flux source area, that higher CO_2values coming from a higher proportion of man-made surfaces and lower CO_2values be accompanied by highly vegetated area within flux source areas. The study site was a net source of CO_2in the atmosphere and the total annual averaged CO₂emissions from the neighborhood of the tower were estimated to be 30.0kg/(m²·a), which was lower than that of 35.5kg/(m²·a) in London. The results would be helpful for understanding the role and function of urban ecosystems in the global carbon cycle, and can provide a reference material for the future urban development planning, in order to reduce the adverse effects of urban sprawl on climate.