Accuracy on measurements and computational results of wind energy reso urce is one of the key factors for successful investment in wind power projects. Meanwhile, the precision of measurements and calculation of the wind energy res ource are primarily dependent on the terrain topography around the wind power fa cilities, distribution of the wind observing stations in the field, performance of observational instruments, and installation of these instruments.  Based on field measurements and analyses of a number of wind power faci lities, in particular over complex terrains, several major measures are proposed on t he quality control of observation data and computational precision. They include 5 principles on distribution of wind measurement stations over complex terra in, reasonable selection and appropriate parameter setting of instruments. Furthermore, quantitative estimations proceeded on systematic errors inherent i n the anemometers, and computational errors possibly arisen from interpolation o f data series for data void points. The computational results based on large qua ntity of field measurements show that: (1) the relative deviations of the import e d anemometers currently in common use are contained between 1.6% and 5.25%, a nd therefore it may produce over 5%, with maximum 13.8%, of the power density er ror of annual mean wind near the top of wind mill. (2) In the zones with monsoon climate, complex topography and significant annual variation of winds, the data observed in the same season or under the same dominant wind direction should be chosen as base data for interpolation, correction of the data void points , and/or for extrapolation of the time series of the data. Otherwise, the relati ve error of the power density of the mean winds may reach as large as 20%～50%.