A set of reduced-order models are considered to determine the variation of the material thermal capacity and thermal conductivity with respect to temperature for a representative hypersonic vehicle structure on a terminal trajectory. The number of thermal degrees of freedom is first reduced by projecting the thermal state of a sample structure into a modal space whose bases are determined using proper orthogonal decomposition. A numerical integration scheme based on the Crank-Nicolson algorithm is used to simulate the thermal state forward in time. Models for the generalized material thermal properties are based on the method of kriging, a least-squares polynomial approximation, and a singular value decomposition approach. The resulting thermal models are compared in terms of accuracy and computational efficiency. The singular value decomposition approach is shown to be the superior overall reduced-order model to capture the variation of thermal properties with temperature when compared to a full-order finite element solution. The effects of varying the number of retained thermal modes and thermal property eigenvectors on the singular value decomposition model are then considered. It is shown that only a few eigenvectors need to be considered to achieve excellent agreement with finite element analysis.