The flash method proposed by Parker et al. in the early sixties is one of the most important experimental procedures to determine the thermal properties of homogeneous materials. Because of the versatility of this method, researchers have attempted to extend its usefulness into the realm of composite material. However, some difficulties arise because of the existence of preferential heat paths in heterogeneous materials, especially in fiber-reinforced composites. In order for experimental flash method results to be meaningful a homogeneous temperature front must exist at the back face of the sample, where the measurements are made. In this work, the parameters that render the radial temperature response of a fiber reinforced composite homogeneous at the back face were investigated. According to the literature three criteria must be met for homogeneity to occur: fiber-to-matrix volume ratio must be high; sample axial dimension must be large compared to radial fiber dimension and the thermal contact between the fibers and the matrix must be high. Since the first two criteria are met by most fiber-reinforced composite samples subject to the flash method, attention was concentrated on the third criterion. An inequality that must be met by the contact conductance term to establish homogeneity is proposed and some sample temperature profiles are presented. / Master of Science