After late frost injury, the physiological and ecological aspects of winter wheat were changed, among which the change of plant height was the most significant. In this research, four plant height components, including plant height, ear length, peduncle length, and penultimate internode length, as well as 15 hyperspectral parameters, namely red edge position and the red edge amplitude, etc were obtained. The change rate plant height components, which has the best correlation with the hyperspectral parameters were selected by correlation analysis, and then were used to establish stepwise regression model. The results showed that only the change rate of plant height was significantly correlated with the hyperspectral parameters in both two experiments in 2013 and 2014. After the two experimental data were combined, the change rates of spike length, peduncle length and the penultimate internode length were also significantly correlated. Comprehensively considering the Adj. R2 and the significance level (Sig. ), it can be seen that the best fitting model is the change rate of ear length, followed by the plant height, the peduncle length and the penultimate internode length. Comparing the RMSE of the model, it can be seen that the change rate of peduncle length got the highest prediction precision. The results of this study provide a good reference for the prediction of wheat plant height by hyperspectral parameters under freezing stress conditions. The results are of great significance to the study of the changes of plant height elements in winter wheat under low temperature stress.