Abstract: The precise calculation of landslide volume seems simple but difficult actually. The Loess Plateau has a few characteristics which include sparse vegetation, high degree exposure of rock and soil, and landslide susceptibility. It has occurred many landslides. Compared with the traditional calculation method, such as remote sensing and 3D laser scanning, digital close-range photogrammetry is little affected by all kinds of factors and has the features of flexible operation, high efficiency and high precision products. This paper uses Dangchuan 2^# loess landslide in Heifangtai tableland as an example. It uses high resolution close-range photogrammetry data to analyze topography change between pre-sliding and post-sliding. Based on the field investigation and carrying out fine mapping of the landslide, this paper realizes the digital landslide. The detailed study consists of landslide influence, landslide area, landslide volume, loose coefficient of landslide volume and the local dry density change. The results are shown the follows.(1) It covers a total area of 105216m², with 217m in north-south direction, and 176m in west-east direction. The average thickness of the landslide is about 20m (slip source area), with sliding 782m, trailing 122m in altitude from the landslide crown to landslide top.(2) According to the characteristics of the sliding and accumulation of each part of the slope, it is divided into the collapse zone, the main sliding zone, the sliding zone, the circulation accumulation area, the shovel winding accumulation area, the extrusion and accumulation area, the two accumulation area and the dust accumulation area.(3) According to high resolution DEM, the sliding volume is 31.72×10⁴m³ and accumulating volume is 49.96×10⁴m³. (4) The overall loose coefficient of the landslide is 1.411. The local dry density has great difference, which has close relationship with the development characteristics and the movement process of the landslide.