The SLA (spaceborne laser altimeter) borne on Gaofen-7 satellite is the first SLA with full-waveform recording for Earth observation. The horizontal and vertical accuracy of laser footprints are the most important technical indicators for a SLA. This paper proposes a method to improve the geolocation accuracy for laser footprints based on waveform simulating and matching. Using airborne lidar point cloud data as in-situ measurements, the laser pointing and ranging information are calculated and further calibrated by matching the simulated waveform with the Gaofen-7 captured waveform of a single footprint and combined treatment of successive footprints correlation coefficient within a laser track. Mountainous study areas (Utah states in USA) were selected. After using the proposed method, the residual of laser footprint elevations in Utah and Wyoming states was improved from 2.54m to 0.27±0.61m with the mean surface slope of approximately 20°, respectively. The results indicate that the proposed method can improve the data accuracy of Gaofen-7 SLA in mountainous areas with the help of in-situ DEM (digital elevation model) data.