Abstract: Abstract: Leaf area index (LAI) is the important vegetation parameter, which is the basis to monitor crop growth and predict crop yield and used widely in many applications. Remote sensing techniques are known to be effective, inexpensive and non-destructive methods for estimating the LAI of crop canopies. And unmanned aerial vehicle (UAV) hyperspectral remote sensing technology as a new method for obtaining agricultural resources, crop growth and other information, can bring new opportunities for LAI dynamic monitoring. As is known to all, spectral characteristic parameters are extremely important for hyperspectral remote sensing in tracking the LAI of crop. However, the research about retrieving the LAI based on UAV hyperspectral remote sensing is rarely reported. And few studies based on spectral characteristic parameters in estimating LAI by using UAV hyperspectral data have been done. In this paper, we carried out the sky-to-ground remote sensing experiments used ASD Field SpecFR Pro 2500 spectroradiometer (ASD) and Cuber UHD185 Firefly imaging spectrometer (UHD185) in winter wheat experimental field. And we acquired two synchronal data including UAV hyperspectral remote sensing data and in-field data for 3 critical growth stages i.e. booting stage, anthesis stage and filling stage. So we retrieved the winter wheat LAI by using the acquired data. Firstly, we regarded the ASD hyperspectral data as the standard spectral data for evaluating UHD185 hyperspectral data. And we evaluated UHD185 data from 3 aspects, i.e. the spectral curve trend, the spectral correlation and the target objects' spectral differences. It demonstrated that the UHD185 spectrums from 458 to 830 nm were reliable data, which could be used to monitor LAI and provide the basis for using UHD185 hyperspectral data in future. Furthermore, we calculated 12 kinds of spectral characteristic parameters, which were 4 red edge parameters including the red edge position (REP), the red edge amplitude (Dr), the red edge area (SDr), and the ratio of the red edge amplitude to the minimum amplitude (Dr/Drmin), and 8 spectral indices including the normalized difference vegetation index (NDVI), the optimized soil-adjusted vegetation index (OSAVI), the triangular vegetation index (TVI), the modified soil-adjusted vegetation index (MSAVI), the modified triangular vegetation index (MTVI1), the modified chlorophyll absorption ratio index 2 (MCARI2), the normalized difference spectral index (NDSI), and the ratio spectral index (RSI). And we analyzed the correlation between the spectral characteristic parameters and the LAI. The results showed that the RSI(494,610) was highly positively correlated with the LAI and better in retrieving the LAI than the others. Meanwhile, a linear regression model created based on the lg(RSI) exhibited that the lg(RSI) and lg(LAI) had the optimum linearity relationship (R2=0.737, n=103). And the predictive lg(LAI) had the high fitting with the measured lg(LAI) (R2=0.783, RMSE=0.127). Therefore, this study confirmed the UAV hyperspectral remote sensing was feasible in estimating winter wheat LAI. It can provide the evidence for developing UAV hyperspectral remote sensing application in agriculture. Moreover, the results also provide a powerful evidence to develop the integration of multiple remote sensing platforms based on satellites, aviation, UAVs and ground objects in precision agriculture.