Wire and cable coverings are potentially a major cause of fire in buildings and other installations. As they need to breach fire walls and are frequently located in vertical ducting, they have significant potential to increase the fire hazard. It is therefore important to understand the ignition and burning characteristics of cables by developing a model capable of predicting their burning behaviour for a range of scenarios. The fire performance of electrical cables is usually dominated by the fire performance of the sheathing materials. The complexity of the problem increases when cable sheathing incorporates fire retardants. One-dimensional pyrolysis models have been constructed for cable sheathing materials, based on milligram-scale and bench-scale test data by comparing the performance of three different software tools (ThermaKin, Comsol Multiphysics and FDS, version 6.0.1).Thermogravimetric analysis and differential scanning calorimetry were conducted on powdered cable coatings to determine the thermal degradation mechanism, the enthalpy of decomposition reactions, and the heat capacities of all apparent species. The emissivity and the in-depth absorption coefficient were determined using reflectance and transmittance measurements, with dispersive and non-dispersive spectrometers and integrating spheres.Bench-scale tests were conducted with a mass loss calorimeter flushed with nitrogen on samples in a horizontal orientation, for comparison with the pyrolysis model of non-flaming decomposition at an external heat flux of 50 kW m -2 . The parameters determined through analysis of the milligram-scale data were used to construct a pyrolysis model that predicted the total mass loss from calorimeter tests in anaerobic conditions. A condensed phase pyrolysis model that accurately predicts in-depth temperature profiles of a solid fuel, and the mass flux of volatiles evolved during degradation of the fuel, is an essential component of a comprehensive fire model, which when coupled to a computational fluid dynamics code can be used to predict the burning processes in a fire scenario. Pyrolysis models vary considerably in complexity based on the assumptions incorporated into the development of the model.