By using the spectral reflectance data ( R λ ) of canopies， seven species of fruit trees were identified during flowering period. Firstly， the identification capacity of six kinds of satellite sensors and four kinds of vegetation index was compared on the basis of resampling the spectral data for six kinds of pre-defined filter function and calculating vegetation index. Then， a BP neural network model for identifying seven species of fruit trees was established on the basis of choosing the best transformation of R λ and optimizing the model parameters. The main conclusions are as follows:(1) the order of the identification capacity of six kinds of satellite sensors from power to weak is: MODIS、ETM+、QUICKBIRD、IKONOS、 HRG、ASTER， (2)among four kinds of vegetation index， the identification capacity of RVI is the most powerful， next is NDVI， and the identification capacity of SAVI or DVI is relatively weak， (3)the identification capacities of RVI and NDVI that are calculated with the reflectances of near-infrared and blue channels of ETM + or MODIS sensor are relatively powerful， (4)among R λ and it's 22 kinds of transformation data， d 1 ［log(1/ R λ )］(derivative gap is set as 9 nm) is the best transformation for setting up BP neural network model.