Vegetation biophysical variables, LAI and FPAR, are the most important terrestrial properties. For acquiring these variables in local scale, the most effective approach is by remote sensing models combined with the groundbased validation. Spectral index model and radiant transmission model are two kinds of key methods. Through the precise radiometric and atmospheric correction, it is possible to obtain the LAI/FPAR products with a high accuracy. There are several factors influencing the accuracy of these products, such as the pixel heterogeneity, vegetation types and growing seasons. LAI and FPAR have the compact relationship with crop yield and they are also the basic variables of many crop growth models. Using them could realize the true yield prediction, especially for estimating the production in the global scale.