The pyrolysis of cellulose was studied by electrically heating in helium single strips (0.75 cm×2.5 cm) of low ash (<0.07%) predried filter paper (≈0.01 cm thick) supported inside a folded wire mesh heating element mounted in a sealed vessel. The samples were rapidly brought to a desired temperature, held there for a desired time and then rapidly cooled. Extents of conversion to volatiles, measured by weighing samples before and after experiments of known duration, were determined as a function of residence time (0.2–75,600 s), final temperature (250–1,000°C), heating rate (400–10,000°C/s), and ambient pressure (0.0005–1 atm). Pyrolysis kinetics was determined by nonisothermal techniques to account for substantial reaction occurring during sample heat-up and cool-down. All of the cellulose was converted to volatiles without char formation, except when an extended heating period at a low temperature (e.g., an hour or longer at 250°C) preceded the pyrolysis at higher temperatures. The latter procedure resulted in char yields of about 2% by weight of the original cellulose. The kinetics data are well described by a single-reaction first-order decomposition model with an activation energy of 33.4 kcal/mole and a frequency factor of 6.79×10 9 s −1. The correlation is slightly improved by use of a multiple-reaction model based on a set of independent parallel first-order reactions represented by a Gaussian distribution of activation energies with a mean of 37.0 kcal/mole and a standard deviation of 1.1 kcal/mole. Selected experiments at reduced pressures (0.0005 atm), elevated heating rates (10,000°C/s), or both resulted in rate constants smaller than those obtained at 400°C/s and 1 atm. The results indicate that the residence time of volatile products within the pyrolyzing cellulose matrix is extremely important in determining conversion. Suggested pathways for cellulose pyrolysis that are consistent with these findings and with much of the pertinent literature involve primary decomposition to an oxygen-rich intermediate (probably levoglucosan) which then participates in three processes to extents depending on experimental conditions: (a) direct escape from the decomposing material into the ambient gas; (b) polymerization, cross-linking and cracking to form char and (c) pyrolysis to smaller volatiles some of which inhibit the char formation in (b) or autocatalyze (c). Accordingly, char is not a primary product, and the yield of char is zero for extremely short residence times of the primary products within the matrix of the decomposing material or for conditions that permit complete inhibition of char formation by secondary pyrolysis products.