Abstract: In order to control the aphid damage in agricultural production efficiently and enhance the timeliness in monitoring and prediction of aphid by remote sensing technology, twenty-six survey points at different aphid damage levels were selected and canopy spectra were collected accordingly by ASD hand hold spectrometer in the filling stage of winter wheat. By comparing and analyzing the characteristics of different aphid damage levels, it was concluded that canopy reflectance gradually decreased within 350～1750 nm, especially in the near infrared region. Further more, the sensitive bands were selected out by relevance analysis in the visible, near-infrared, and short-wave infrared region, respectively, and the aphid hyperspectral vegetation index (ADHI) was established based on the most sensitive bands. Finally, the regressive model for ADHI and aphid damage levels was established, and the results clearly showed the potential of retrieval model (R2=0.839, n=26) for quantifying aphid damage levels in the filling stage of winter wheat. The results can provide a theoretical basis for identifying aphid and monitoring aphid damage levels.