There is significant controversy in the reduction kinetics of chemical-looping combustion (CLC) between NiO and CH4. We propose an application of a model-based framework to improve the quality of CLC experiments with respect to model discrimination and parameter estimation. First, optimal experiments are designed and executed to reject inadequate models and to determine a true model structure for the reaction kinetics of the CH4-NiO system. Then, kinetics with statistical significance is estimated from experiments aimed at reducing parameter uncertainty. To maximize the observability of the NiO reduction reactions, fixed bed experiments should exhibit a peak separation of the concentration profiles, an initial high methane slip, and low overall CO2 selectivity. Several case studies are presented to check the adequacy of the recommended model and evaluate its predictive ability and extrapolation capabilities. The model resulting from this work is validated and suitable for application in process design and optimization. (c) 2016 American Institute of Chemical Engineers