Moisture and Calorific value, which are two of the most important properties of straw for energy conversion process, were predicted by near infrared spectroscopy (NIRS) technique combining the use of LOCAL algorithm. Firstly the samples were evenly divided into 3 subsets according to the chemical values, named low, mid and high concentration respectively, to build global PLS calibrations. Standard errors of cross validation (SECV) of the three calibrations based on subsets were almost lower than that of calibration based on the whole samples, suggesting that the variation of moisture and calorific value affect the accuracy of NIRS calibrations. Then LOCAL algorithm was introduced to near infrared spectroscopy analysis for rapid measurement of the moisture content and calorific value of straw samples. By the use of LOCAL algorithm, which patented by Shenk to reasonably select a group of samples for each sample to build partial least square regression (PLS) calibration, the prediction accuracy were improved compared to the PLS and MPLS models for both moisture and calorific value of straw. It was therefore concluded that LOCAL algorithm applied in quantitative analysis of straw was able to improve the prediction accuracy.