Chlorophyll is the main pigment in photosynthesis of crops to synthesize organic matter.Real-time monitoring of chlorophyll content in flue-cured tobacco leaf plays an important guiding role in tracking nitrogen nutrition status of tobacco and distinguishing the maturity of tobacco leaves. Based on the analysis of spectral characteristics of flue-cured tobacco leaves，the measured hyperspectral data and chlorophyll relative content(SPAD)of tobacco leaves under different nitrogen levels were used as data sources.The raw spectral data were preprocessed by first derivative(1st Der)，multiple scattering correction(MSC)，standard normal variable(SNV)and Savitzky-Golay smoothing(SG). First，the successive projection algorithm(SPA)was used to select characteristic wavelengths under the different preprocessing conditions，and then the spectral reflectance of the extracted characteristic wavelengths was used as input variable of back propagation neural network(BPNN)，random forests(RF)and support vector machine(SVM)modeling.Three machine learning methods were used to construct estimation models of chlorophyll content in inverted three leaves of flue-cured tobacco，respectively.The performance of each machine learning model was evaluated and compared using the coefficient of determination(R2)，root mean square error(RMSE)and mean absolute error(MAE). The results showed that:Compared with the models trained by the three machine learning methods，the RF model achieved the best prediction result.In this study，the raw spectral data of flue-cured tobacco leaves pretreated by MSC and SNV were used as the input variables of the model，and then the model was established by RF algorithm which had obtained the best prediction results.The three models were as follows:MSC-SPA-RF(R2=0.96，RMSE=1.15，MAE=0.94)and SNV-SPA-RF(R2=0.96，RMSE=1.14，MAE = 0.94). It is feasible to estimate chlorophyll content of flue-cured tobacco leaves using hyperspectral data based on machine learning.This provides a theoretical basis for monitoring real-time and accurate change of chlorophyll content in flue-cured tobacco leaf and for reasonable and scientific field management.