Abstract: Presently, the eddy covariance measurement has been widely adopted. As the eddy covariance method is built on certain assumptions, some corrections of turbulent flux calculation and quality assessment of the flux results are necessary. This article analyzes the difference of turbulent flux results before and after quality control, with the data at Litang station from mid-July to mid-August in year 2006 as an example. The results at Litang station show that the de-spiking and sonic temperature corrections have a little effect on the sensible and latent heat fluxes, the coordinate rotation correction significantly influences the momentum flux, and because of the larger latent heat flux, the WPL correction have an inconspicuous influence on the latent heat flux. The conclusion of quality assessment shows that 68.2% of the momentum flux data, 60.6% of the sensible heat flux data, and 63.3% of the latent heat flux data are good.