Near infrared spectroscopy (NIRS) has been broadly applied as a quick assay for biological component and property analysis. However, NIRS remains unavailable for in-depth analysis of biomass digestibility in plants. In this study, NIRS was used to determine biomass enzymatic digestibility using 199 Miscanthus samples, which represents a rich germplasm resource and provides for a stable calibration model. The intensive evaluation indicates that the calibration and validation sets are comparable. Using the modified partial least squares method, seven optimal equations were generated with high determination coefficient on calibration (R-2) at 0.75-0.89, cross-validation (R(2)cv) at 0.69-0.87, and the ratio performance deviation (RPD) at 1.80-2.74, which provide multiple options for NIRS prediction of biomass digestibility under different pretreatments. As biomass digestibility is a crucial parameter for biofuel processing, NIRS is a powerful tool for the high-throughput screening of biomass samples in plants. (C) 2012 Elsevier Ltd. All rights reserved.