Recent experimental work has shown that methanol to olefin (MTO) catalysis on microporous solid acids proceeds through a hydrocarbon pool mechanism with methylbenzenes frequently acting as the most important reaction centers. Other recent experimental evidence suggests that side-chain methylation is more important than an alternative paring (ring contraction-expansion) mechanism. The present investigation uses density functional theory B3LYP/cc-pVTZ//B3LYP/6-311G* and G3(MP2) calculations to model many of the features of the side-chain mechanism. We first calculated at the G3(MP2) level the heats of formation of 43 neutral alkybenzenes to predict the thermodynamics for methylation reactions. The G3(MP2) results predict that sequential methylation of benzene rings with fewer than four methyl groups will preferentially occur on the ring, resulting in the series toluene, 1,3-dimethylbenzene, 1,2,4-trimethylbenzene, and 1,2,4,5-tetramethylbenzene. With the addition of another methyl group side-chain methylation becomes preferred, with 1-ethyl-2,4,5-tetramethylbenzene predicted to be more stable than pentamethylbenzene by 0.7 kcal/mol. We modeled the entire gas-phase side-chain reaction mechanism at the l33LYP/cc-pVTZ//B3LYP/6-311G* level, using p-xylene, 1, 2,3,5-tetramethylbenzene, and hexamethylbenzene as reaction centers and following the reaction to the point of producing both ethylene and propene. B3LYP/6-311G* analytical frequencies were calculated in order to obtain the data needed for the prediction of enthalpies. For comparison, G3(MP2) enthalpies were also calculated for the mechanism based on rho-xylene only. We also used a zeolite cluster model to more accurately describe the relative energetics of the reaction for the entire hexamethylbenzene mechanism and parts of the rho-xylene mechanism. These calculations place the side-chain mechanism on a much stronger foundation and reproduce experimental structure-reactivity and structure-selectivity data for the methylbenzene hydrocarbon pool.