The thermal conductivities of as-sprayed yttria-stabilized zirconia thermal barrier coating prepared by atmospheric plasma spraying at different temperatures are investigated based on quantitative microstructural analysis. Multiple linear regression is used to develop quantitative models which describe the relationship between multiple elements such as porosity, grain boundary density, monoclinic phase content, temperature and thermal conductivity. Results reveal that the thermal conductivity of the coating is mainly determined by the porosity and grain boundary density below 300 °C and by the monoclinic phase content above 800 °C. Furthermore, based on the significance testing analysis, the confidence interval under a confidence level of 95% at different temperatures enables researchers to predict the thermal conductivity based on microstructural information.