Five kinds of straw including cotton straw,wheat straw,corn straw,rape straw,rice straw collected by carbonization processing were used as straw biochar samples.Calorific value were measured with the C2000 heat meter and infrared spectrum were collected with the VERTEX type 70 Fourier infrared spectrometer.Leverage and residual was used to eliminate abnormal samples.SPXY was used to divide the sample set.TQ Analyst 8.3 was used to preprocess the spectra.The PLS model was established to predict the gross and net calorific value of straw biochar.The results showed that the multiple scatter correction (MSC) in combination with first derivative was the best model of preprocessing gross calorific value,with correlation coefficient of 0.909,verify the root mean square error of 614 J/g,and relative standard deviation of 2.61%.Multiple scatter correction (MSC) and first derivative in combination with the S-G smoothing was the best model of preprocessing net calorific value,with correlation coefficient of 0.873,verify the root mean square error of 707 J/g,and relative standard deviation of 3.12%.It is indicated that using FT-MIR to predict calorific value of straw biochar is feasible.