For the stability analysis of rock slope, it is very critical to obtain the spatial geometric characteristics of the structural surfaces of the rock mass accurately and effectively. As for a high-steep rock slope of an iron ore mine, in order to solve the problems of inefficiency and high risk of traditional manual geological survey, the geological survey and stability evaluation of the slope was carried out by adopting unmanned aerial vehicle digital photogrammetry (UAV-DP) technology. Firstly, a large number of high-resolution images of the slope were obtained by UAV-DP. Then, the structure from motion (SFM) method was used to construct the fine 3D point cloud model of the slope, which was subjected to coplanarity detection and K-means clustering for identifying the structural surfaces. Finally, the stability and failure model of the slope cut by the structural surfaces are analyzed by using the stereo-projection and discrete element methods. The research results show that the error between UAV-DP and manual measurement is within the acceptable range, which demonstrates the reliability of UAV-DP used in the geological investigation. Furthermore, the stability state and failure model of the slope is also consistent well with the field observation.