• Optimised process parameters for thermal recycling of thermoset carbon fibre reinforced polymer is presented. • Minimisation of carbon fibre degradation and maximisation of matrix removal is investigated. • Modelled results are validated with experimental and analytical results. • Improved integrity of recovered carbon fibre for achieving closed-loop recycling is discussed. A mathematical prediction model for identifying the most suitable process parameters for thermal recycling of thermoset carbon fibre reinforced polymer (TS-CFRP) composites is presented here. The process parameters of heating rate, maximum temperature and isothermal dwell time at the maximum temperature for recovery of carbon fibres depend on factors like the size of furnace and composite sample, composition and constituents of composites. The model considers such factors to give the most suitable process parameters so that the recovered carbon fibre has minimum degradation yet with maximum matrix removal. Modelled results on heat transfer, material conversion, radiation view factor, and predicted process parameters were verified with the experimental results. The model can provide a good prediction of the furnace process parameters and final matrix mass loss distribution by layer with an error percentage of five per cent. Using this prediction model with the TS-CFRP thermal recycling processes can assure conserving the original fibre value for closed loop recycling of carbon fibre.