One hundred fifty Leymus chinensis samples with different growth stage, areas, and preparing method (oven-drying and shading natural dry), were selected to study the potential of determination of crude protein(CP), neutral detergent fiber(NDF)and acid detergent fiber(ADF)in the present research. The quality parameters of Leymus chinensis were firstly predicted using the near infrared reflectance spectroscopy in China. The three models were validated by cross-validation and external-validation. The results indicated that the NIRS models of Leymus chinensis quality prediction highly accessed the precision of chemical analysis. The coefficient of correlation of cross-validation of crude protein, neutral detergent fiber and acid detergent fiber were 0.963 7, 0.959 4 and 0.947 9, and the RMSECV of the three models were 1.41%, 1.27% and 1.27%, respectively; the correlation coefficients of external-validation were 0.965, 0.956 and 0.953, and all the ratios of standard deviation to root mean square error of prediction were higher than 3. Thus it can be testified that using NIRS analysis can rapidly and accurately determine the quality of Leymus chinensis. This method is of great significance for quick analysis of the trait of Leymus chinensis production and screening of breeding materials in Leymus spp. research of China.