The longitudinal load in rail caused by thermal expansion must be regularly monitored in order to avoid buckling or rail fracture. Different methods of monitoring with benefits and drawbacks are used or suggested. In this paper one of the proposed methods is investigated by a full-scale experiment. The aim is to measure the change in wavelength of the bending wave caused by the longitudinal load. In contrast to other vibration methods, this method does not require knowledge of the boundary conditions. However, it requires very accurate measurements, advanced finite element (FE) calculations, and sophisticated data analyses. The full-scale experiment shows that this is a method with potential. On the basis of the results of the full-scale experiment the required accuracy of the different steps in the method are clarified. Influence of measurement accuracy, loosened clamps at the sleepers, FE mesh size, degree of wear of the rail, and inaccuracy in the material parameters is considered.