The wheel–rail contact problem plays an important role in the simulation methods used to solve railway dynamics problems. As a consequence, many different mathematical models have been developed to calculate wheel–rail contact forces. However, most of them tackle this problem purely from a theoretical point of view and need to be experimentally validated. Such validation could also reveal the influence of certain parameters not taken into account in the mathematical developments. This paper presents the steps followed in building a scaled test-bench to experimentally characterise the wheel–rail contact problem. The results of the longitudinal contact force as a function of the longitudinal creepage are obtained and the divergences with respect to Kalker's simplified theory are analysed. The influence of lateral creepage, angular velocity and certain contaminants such as cutting fluid or high positive friction modifier is also discussed.