This paper investigates near infra-red spectroscopy (NIRS) as an indirect and rapid method to assess the biochemical methane potential (BMP) of meadow grasses. Additionally analytical methods usually associated with forage analysis, namely, the neutral detergent fibre assay (NDF), and the in-vitro organic matter digestibility assay (IVOMD), were also tested on the meadow grass samples and the applicability of the models in predicting the BMP was studied. Based on these, regression models were obtained using the partial least squares (PLS) method. Various data pre-treatments were also applied to improve the models. Compared to the models based on the NDF and IVOMD predictions of BMP, the model based on the NIRS prediction of BMP gave the best results. This model, with data pre-processed by the mean normalisation method, had an R(2) value of 0.69, a root mean square error of prediction (RMSEP) of 37.4 and a residual prediction deviation (RPD) of 1.75. (C) 2011 Elsevier Ltd. All rights reserved.