Biochemical methane potential (BMP) is the most relevant test measured to characterize a waste in order to produce biogas (methane + carbon dioxide). This test is based on a fermentation process which are time consuming, about 30 days or more for some wastes. The issues is to reduce the time of analyse to one day. The BMP value depends only on the quantity and the quality of organic matter. Near infrared spectroscopy can analysed the organic matter of the waste in short time. It's the first time NIRS is used to predict the BMP value of solid waste. NIRS has the advantages to be very fast and applicable directly on the solid waste after being freeze-dried and grinded. BMP value and NIR spectra were measured on 68 samples of municipal solid wastes. Pre-treatment (SNV, Detrend) and partial least square regression (PLS) have been used to calibrate the model of prediction. Results demonstrate the feasibility of the NIRS to predict the BMP value of municipal solid waste (r²: 0.76 ; Standard error of prediction: 28 ml CH4.g-1 Volatil Solid. The observation of the regression coefficient could be used to approach the biodegradability of the waste, showing which organic fraction is more biodegradable than an other. ). This results is very good for an industrial application.