The potential of near infrared spectroscopy (NIRS) for determining fermentable substance and also ethanol yield in wheat, rye and triticale grains as a feedstock for fuel ethanol production was investigated. The total sample set contained 480 samples of 10 types of wheat, 24 types of triticale and 6 types of rye, which were grown on 10 locations in Germany from 2006 to 2008. Samples were scanned by NIRS as whole or ground grains, and their reference values regressed against different spectral transformations by partial least squares regression (PLS-1). Ground grains were scanned on a dispersive NIR-Spectrometer. Whole grains were scanned on a diode array NIR-Spectrometer. Principal Component Analysis (PCA) revealed that samples could be classified by crop season, growth location and species. It is shown that near infrared spectroscopy is an appropriate and useful tool for prediction of biofuel yield for both industry and research. (C) 2010 Elsevier Ltd. All rights reserved.