Cellulose is the major constituent of most plants of interest as renewable sources of energy and is the most extensively studied form of biomass or biomass constituent. Predicting the mass loss and product yields when cellulose is subjected to increased temperature represents a fundamental problem in the thermal release of biomass energy. Unfortunately, at this time, there is no internally consistent model of cellulose pyrolysis that can organize the varied experimental data now available or provide a guide for additional experiments. Here, we present a model of direct cellulose pyrolysis using a multistage decay scheme that we first presented in the IJQC in 1984. This decay scheme can, with the help of an inverse method of assigning reaction rates, provide a reasonable account of the direct fast pyrolysis yield measurements. The model ลฝ . is suggestive of dissociation states of d-glucose C H O , the fundamental cellulose 6 10 5 monomer. The model raises the question as to whether quantum chemistry could now provide the dissociation energies for the principal breakup modes of glucose into C , C , 1 2 C , C , and C compounds. These calculations would help in achieving a more 3 4 5 fundamental description of volatile generation from cellulose pyrolysis and could serve as a guide for treating hemicellulose and lignin, the other major biomass constituents. Such advances could lead to the development of a predictive science of biomass pyrolysis that would facilitate the design of liquifiers and gasifiers based upon renewable feedstocks.