Abstract High-strength steels, despite their excellent mechanical properties in normal conditions, can be susceptible to hydrogen embrittlement. Due to the service loads or residual stresses, hydrogen migrates within the component and accumulates in the regions where the highest tensile hydrostatic stress occurs. As a consequence, component brittle failure can occur even if the initial or mean hydrogen concentration is lower than the critical value. The availability of models predicting the hydrogen diffusion within the component is a crucial task for the design. Several diffusive models have been presented in the literature and some general-purpose finite element codes have already implemented some of them. However, the validation of those models is still an open issue due to the difficulty in performing accurate local measurements of the hydrogen concentration. This study deals with the design of a test potentially able to validate hydrogen migration models. In the test, a four-point bending configuration is applied to a properly shaped hourglass specimen, previously charged with hydrogen, extracted from thin high-strength steel sheets. The specimen geometry and the loading configuration were designed to obtain a central region in which the stress and strain field is uniform in plane and exhibits a quasi-uniform gradient in the thickness direction. As a consequence, it is expected a large enough central region of the specimen in which the Hydrogen can migrate only in the thickness direction during the typical duration of the test. The local hydrogen concentration is evaluated by measuring the flux leaving the tensile surface of the specimen by a solid-state hydrogen sensor.