During SAE BRASIL/96 a paper was presented showing that it was possible to represent the influence of the coefficient of friction in the stamping process using numerical simulation and some results were shown where it became clear that the correct definition of the friction conditions is very important for the results reliability. It was also shown that the current practice of assuming a typical value for the coefficient of friction, independent of the lubrication conditions, can lead to imprecise results. Along the drawing operation, the blank contacts different regions of the die and for each region there are different contact conditions (contact pressure, sliding velocity, etc.). These differences can lead to friction conditions where the coefficient depends on the contact variables and the Coulomb friction model is not applicable. However, controlling those contact variables, it is possible to develop a methodology to adapt the use of simulation Finite Element (FE) codes with the Coulomb model implemented to represent a different friction model that can consider the local contact conditions.Using a simple friction test device, a procedure was developed to define the coefficient of friction under different contact conditions, in order to identify the values to be used in stamping simulations at the different regions of contact between the blank and the die surfaces. The coefficients of friction experimentally determined were applied in the process simulation of a stamped part that was also analyzed assuming an average uniform value for the coefficient. The comparison of both conditions results indicated that the use of different coefficients of friction depending on the contact conditions leads to small differences on the results when compared with the ones obtained using a uniform coefficient. It was also verified that the experimental procedure needs some improvements in order to allow the characterization of all the contact conditions between the blank and the die surfaces.