Advances in instrumentation, notably the eddy covariance(EC) technique, offer a tool for the direct measurement of representative flux data in urban areas.The complexity of the urban underlying surface and limitations of the EC system cause the obtained surface fluxes to not be true values.If no correction is applied, the error will be greater than 10%.The applicability of the EC technique, quality assessment of the EC system, and processing of its data, in urban Beijing, were analyzed using the CO₂ flux data and energy measured at seven levels of the 325-m meteorology tower in the city.Analysis of the data from the tower showed that the best time period to calculate the CO₂ flux for the urban underlying surface is 30 minutes.The secondary coordinate rotation method is superior to plane fitting.After frequency response correction, the sensible heat flux increased by 5.21%, and the latent heat flux and CO₂ flux were increased by 9.42%.Liu's method(Liu, 2005) was found to be better than the Webb E K, Pearman G I, and Leuning R's(WPL) method(Webb et al., 1980) in the air density pulse correction step.The turbulent spectral checks of the data quality evaluation were satisfactory.After rigorous data screening, the final result showed that approximately 79% of the flux data were good.