We present a novel method for calculating degradation kinetics in polymers. Our calculations directly use the dissociation energy of chemical bonds in a polymer chain to predict weight loss as a function of time and temperature in an Arrhenius-type activation function. The novelty lies in quantifying the thermal energy term for skeletal bonds in the chain backbone in the activation function that initiates the bond fission process and that also quantifies the pre-exponential rate term. Our method allows prediction of TGA experiments with any time-temperature profile directly from the polymer structure using tools such as quantum mechanics simulations for bond dissociation energy. The model is demonstrated by application to a number of synthetic polymers with different temperature ramp rates. Application of the model to protein polymers shows significant differences from synthetic example polymers, since the synthetics use a single average dissociation energy, whereas proteins seem to degrade sequentially with the individual skeletal bond energies. (C) 2009 Elsevier Ltd. All rights reserved.