The feasibility of fuel property analysis of the biomass fuel by laser induced breakdown spectroscopy (LIBS) has been demonstrated in a previous study. As a further move to apply this quick method to industrial application for direct monitoring of the woody biomass pellets quality, effects of four key parameters during the pelletizing process on the LIBS spectral data have been investigated before applying LIBS directly to the woody biomass pellet. Partial Least Squares (PLS) models were built to evaluate the performance of the LIBS fuel property analyses directly on the wood pellets supplied by different manufacturers, which were produced on different pellet mills, with different pelletization technologies and under various pelletizing conditions. The results showed satisfactory performance using LIBS technique for direct determination of the fuel property of wood pellets without further sample preparation. The coefficient of determination (R2) of calibration models were all above 0.98. The root mean square error of prediction (RMSEP) of the gross calorific value (GCV), volatile matter (VM) and ash content (AC) were 0.09 MJ/kg, 0.52% and 0.15% respectively, while the average relative error (ARE) were 0.39%, 0.54% and 8.52% respectively, and the average standard deviation (ASD) were 0.14 MJ/kg, 0.52% and 0.18% respectively. (C) 2020 Elsevier Ltd. All rights reserved.