When relatively hard surfaces that make asperity contact slide against one another at speeds on the order of a mls or more very high local temperatures can be generated. If the sliding conditions are of sufficient severity and duration, thermal distortions at local or component levels can occur. Overall thermal deformations are primarily determined by thermal gradients. We have found that during the early parts of a sliding interval, all heat input is confined to small volumes at individual asperities which form "hot mounds" surrounded by much larger cool regions where there is little or no temperature change. Thermal distortions are essentially non-existent and overall component level interactions, including thermal softening, are not much different from isothermal. Eventually, more of the surface and near surface regions are heated and the possibility of component level thermal deformation increases.The modeling of such problems with reasonable fidelity using modern numerical approaches (e.g., finite or boundary element methods) remains a challenge. Asperity level up to component level analysis must be performed simultaneously. When the problem is isothermal, the macro level interactions can be handled by replacing the rough surface by a distributed contact compliance while retaining smooth surface geometry in the analyses. However, with thermomechanical problems, the assumption of smooth surfaces, however implemented, leads to thermomechanical deformations and macroscopic contact pressure distributions that are not found in practice.In this paper, we present some examples of these phenomena both via experimental observations, including shapes of wear tracks, and coupled thermomechanical finite element analyses of smooth and rough surface contacts. Neither view provides a complete picture of the interactions. We describe the current state of the art as we understand it and discuss approaches being taken to model the full problem.