In the present study, ant colony optimization algorithm is used for an inverse heat transfer problem of parameter estimation. Temperature dependent thermal conductivity and specific heat are estimated simultaneously. Performance of the present algorithm is first analyzed by using the temperature data obtained from the solution of direct problem. Effect of measurement error on the property estimation is then analyzed. Accuracy of estimated properties is found to be good in both the cases. Experiments are then conducted on a cotton duck fabric exposed to 40kW/m2 radiant heat flux using a bench top test according to ISO 6942. Temporal variation of sensor temperature is used to estimate the thermal conductivity and specific heat of the fabric sample. The estimated values are in agreement with the results available in the literature. Property values estimated in the study are included in a coupled conduction-radiation heat transfer model for fabrics. Performance of the fabric sample is further analyzed with the numerical model for high heat fluxes. The simulation results compare quite well with experimental data at 80kW/m2 radiant heat and 70kW/m2 flame exposures. Flame exposure experiment is conducted according to ISO 9151. Good agreement between the numerical and experimental results are found for both the exposures.