The enzymatic hydrolysis of cellulose by cellulases is one of the major steps in the production of ethanol from lignocellulosics. However, cellulosic biomass is not particularly susceptible to enzymatic attack and crystallinity of the substrates is one of the key properties that determine the hydrolysis rates. In this work, by quantifying the respective contributions of amorphous and crystalline cellulose to the X-ray diffraction spectra of cellulose with intermediate degrees of crystallinity, a new method to obtain consistent crystallinity index values was developed. Multivariate statistical analysis was applied to spectra obtained from phosphoric acid pretreated cellulose samples of various intermediate (but undetermined) crystallinity indices to reduce their dimensionality. The crystallinity indices obtained were found to be linearly related to the enzymatic hydrolysis rates. The method was validated by predicting the degree of crystallinity of samples containing various ratios of microcrystalline cellulose and amorphous cellulose, both of known crystallinity indices. Dimensionality reduction of the spectra was also used to predict the enzymatic hydrolysis rates of various cellulose samples from X-ray data. The method developed in this work could be generalized to accurately assess the degree of crystallinity for a wide range of varieties of cellulose. (C) 2010 Elsevier Ltd. All rights reserved.