Hot forming die quenching is used by the automotive industry to produce ultra high strength steel parts with a high strength-to-weight ratio. Crash-resistant structural parts with distributed properties can be obtained by controlling the local cooling rate during the quenching step using heated dies. This procedure requires detailed spatial knowledge of the heat transfer coefficient at the blank/die interface. Hot stamping experiments were conducted on Usibor (R) 1500P boron steel blanks to investigate how pressure and blank and die temperatures influenced the heat transfer coefficient, which was inferred using inverse heat conduction analysis. The heat transfer coefficient was found to vary throughout the experiment with deformation of the surface roughness peaks and evolution of the blank and die temperatures. Whereas the heat transfer coefficient at the beginning of the stamping process increases with the initial die temperature, it converges to a value that depends only on applied pressure. Moreover, the experimental heat transfer coefficient for zero pressure was found to match the air gap conductance predicted by semi-empirical models, but the pressure-dependent component was lower than the model-predicted solid contact conductance. (C) 2013 Elsevier Ltd. All rights reserved.