In order to improve the detection rate of wind shear in coherent lidar, a new method to detect wind shear in frequency domain is proposed in this paper.Firstly,the Fourier transform is carried out by intercepting the echo signal at the detection range of the lidar target.Secondly,the frequency component of the echo signal is weighted by its magnitude and the spectrum is reconstructed with the data.Thirdly,the peak frequency is estimated on the bas is of the smoothed reconstructed spectrum.Then, in order to identify the wind shear in the frequency domain,the local frequency difference is calculated and compared with the wind shear threshold.The error of wind shear detection algorithm is evalu ated by using the experimental data of rotating machine, and the spectral estimation error of the algorithm is 0.64MHz.The lidar wind sh ear detection experiment based on small aircraft was designed to verify the detection rate of the algorithm,and the small aircra ft flight report was used as the wind shear criterion to judge the lidar detection results.Through the Chaoyang Airport experiment,t he wind shear detection rate is 84.62%,which shows that the algorithm is of great value to improve the lidar wind shear detec tion rate.