Abstract: A high spectral resolution lidar(HSRL) for measurement of low-stratosphere temperature by scanning Fabry-Perot interferometer (FPI) is proposed and demonstrated. The transmission of Rayleigh backscatter through the FPI is obtained by scanning the cavity spacing of the FPI, and then fitted to Gaussian function using the nonlinear fitting algorithm. Temperature is calculated from the fitted bandwidth of the measured transmission. To reduce systematic error due to frequency instability of the laser, another solid FPI is incorporated into the optical receiver to monitor the frequency drift, which compensates in the data processing. The statistical error is calculated based on a maximum likelihood estimator, which is less than 1.9 K/9.8 K below 30 km/50 km. In the comparison experiment, the max temperature deviation between the high spectral resolution lidar (HSRL) and radiosonde is 4.7 K from 18 km to 36 km, and it is 2.7 K between the HSRL and Rayleigh integration lidar (RIL) from 27 km to 34 km. The temperature profile from Rayleigh integration lidar deviates from the results from HSRL and radiosonde obviously from 15 km to 27 km, with a max deviation of 22.8 K, which may due to the aerosol contamination.