Abstract: Objective To understand the effect of different temperatures on the SPAD value of cotton leaves,the SPAD value of leaf was retrieved by hyperspectral data during the blossoming and boll-forming period.MethodThe statistical methods,such as the correlation analysis and linear regression,were used to analyze the relationship between SPAD value,original spectra,first-order differential spectra and hyperspectral parameters.ResultIt was shown that the SPAD and original spectral reflectivity of the leaf would decrease with the rising temperature.The SPAD value was correlated to the original spectra,first-order differential spectra and hyperspectral parameters.After the overall comparison of correlation coefficient,adjusted R2 and the optimization of mean relative error,the optimal estimation mode for the SPAD value,original spectra,first-order differential spectra and hyperspectral parameters were built,with the correlation coefficient being 0.81,0.857 and 0.833,the adjusted R2 being 0.747,0.844 and 0.824 as well as the mean relative error being 9.12;,5.78;and 7.72;.The model built through the combination of the first-order differential spectra 671 nm and 683 nm(Y=50.487 X683-131.617 X671+36.777)can achieve the highest accuracy in predicting the SPAD value,followed by the hyperspectral parameter and original spectra.ConclusionAs shown by the experimental results,the hyperspectral sensing information can provide the theoretical basis for predicting the SPAD value for the cotton leaf during the blossoming and boll-forming period.