Aero-engine gas turbine performance and efficiency can be improved through the application of compliant shaft seal types to certain sealing locations within the secondary air system. Leaf seals offer better performance than traditional labyrinth seals, giving lower leakage flows at design duties. However, for aero-engine applications, seal designs must be able to cope with relatively large off-design seal closures and closure uncertainties. The two-way coupling between temperatures of seal components and seal closures, through the frictional heat generated at the leaf–rotor interface when in contact, represents an important challenge for leaf seal analysis and design. This coupling can lead to leaf wear and loss, rotor overheating, and possibly to unstable sealing system behavior (thermal runaway). In this paper, we use computational fluid dynamics (CFD), finite element (FE) thermal analysis, and experimental data to characterize the thermal behavior of leaf seals. This sets the basis for a study of the coupled thermomechanical behavior. CFD is used to understand the fluid-mechanics of a leaf pack. The leaf seal tested at the Oxford Osney Laboratory is used for the study. Simulations for four seal axial Reynolds number are conducted; for each value of the Reynolds number, leaf tip-rotor contact, and clearance are considered. Distribution of mass flow within the leaf pack, distribution of heat transfer coefficient (HTC) at the leaf surface, and swirl velocity pick-up across the pack predicted using CFD are discussed. The experimental data obtained from the Oxford rig is used to develop a set of thermal boundary conditions for the leaf pack. An FE thermal model of the rig is devised, informed by the aforementioned CFD study. Four experiments are simulated; thermal boundary conditions are calibrated to match the predicted metal temperatures to those measured on the rig. A sensitivity analysis of the rotor temperature predictions to the heat transfer assumptions is carried out. The calibrated set of thermal boundary conditions is shown to accurately predict the measured rotor temperatures.