Experimental or theoretical modeling of the thermal strength of road materials and structures is usually associated with replacing the study of the effect we are interested in nature (prototype) with the analysis of the solutions of the corresponding mathematical equations or with the consideration of a similar phenomenon on a smaller or larger scale experimental model in special laboratory or experimental conditions with full or partial observance of the correspondence between the physical properties of the objects of nature and the model. The main content of modeling is that, based on the results of mathematical or experimental experiments, it is possible to draw conclusions about the behavior of the system in natural conditions. As a rule, modeling is based on consideration of geometrically or physically similar deformation phenomena of structures. At the same time, it is considered that two bodies are geometrically similar, if the ratio of all corresponding lengths is the same, that is, equal to the coefficient of similarity or scale. Two phenomena are physically similar if the given characteristics of one of them can be used to determine the characteristics of the other with the transition from one system of units to another. As a result of establishing a system of parameters defining a selected class of phenomena, conditions for the similarity of two phenomena can be established. The task of determining similarity conditions for road materials and structures becomes more complicated when they are operated in conditions of variable temperatures, and the number of parameters for which it is necessary to calculate scale factors increases. In this work, they are determined on the basis of the theory of thermoelasticity for geometric characteristics, elasticity parameters (Young's modulus, Poisson's ratio), thermal parameters (linear thermal expansion coefficient, thermal conductivity coefficient). Similarity conditions are defined for the thermal displacement, thermal stress, and temperature functions. Concrete examples also show that methods of similarity theory can be a tool for solving applied problems.