Understanding the thermal decomposition of fuels and estimating their kinetic parameters are essential for simulating chemical reactions in numerical models. In this work, 2-step, 3-step, 4-step, and 5-step kinetic mechanisms for bitumen combustion were developed. The kinetic parameters were optimized via inverse modelling (genetic algorithm) by coupling thermogravimetry (TG) and differential thermogravimetry (DTG), conducted at 5, 10, 20, and 40 °C min−1 under nitrogen and air atmospheres. A 3-step mechanism that includes competing pyrolysis and oxidation reactions was identified as the simplest mechanism able to appropriately simulate all TG experiments, thus avoiding the need for more complex mechanisms. A unique set of kinetic parameters was found by averaging all the parameters optimized at different heating rates and atmospheres, resulting in an average error of 6% when compared with experimental data. This is the first time that averaged optimized parameters were employed, providing similar results as optimizing against all experiments at once. Differential scanning calorimetry experiments were used to calculate the heat of pyrolysis and oxidation, and showed that char oxidation provided the highest energy release, whereas bitumen and asphaltene oxidation represented a 30–110 times lower heat of reaction. This is the first time that thermogravimetry and differential scanning calorimetry experiments were used to optimize kinetic parameters for bitumen combustion.