To prevent soil and water loss, remote sensing and GIS are often applied to soil and water loss investigation. This paper summarizes the techniques estimating vegetation factor by remote sensing in soil and water loss investigation. Vegetation fractional cover is estimated from NDVI using the dimidiate pixel model, by taking Kaixian County of Chongqing City for an example. Better results have been achieved using the technique. By combining the results with newly soil and water loss investigation by remote sensing in Chongqing City, flaws found in the investigation and the relevant suggestions are presented. The evaluation indexes for soil and water loss investigation by remote sensing are also discussed.