In October of 2009 Tampere University of Technology (TUT) and the Finnish Rail Administration (now part of Finnish Transport Agency) conducted a full-scale railway embankment failure experiment in Salo, Finland. Goals of the test were to collect extensive monitoring data for further development of different stability calculation methods and to test the technical suitability of different methods for monitoring the stability of embankments.A new, small railway embankment was built in place of an old, de-commissioned blind track on a clayey soil. Four steel frameworks, each 12 meters long, were used to simulate short railway cars with bogie carriages. The frameworks were loaded with modified shipping containers which were gradually filled with sand. The test area was extensively instrumented with 40 pore pressure gauges, 9 inclinometer tubes, 2 total stations monitoring a total of 27 prisms, 9 earth pressure gauges, 3 settlement tubes and 76 slip surface measuring pipes. In addition, the weight of the containers was measured with strain gauges on the frameworks. Acceleration gauges were used to measure the tilt angle of the containers. The loading of the containers took place during two days. During the initial loading the displacements and pore pressures increased fairly linearly. The rates of displacement and pore pressure increase began to accelerate exponentially at about 2.5 hours before collapse, during the final stages of loading. The final train load was 87 kPa, which is equivalent to a 218 kN/m line load. The resulting landslip (about 50 m wide along the embankment) had a distinctively wedge-shaped cross-section. A clearly defined slip surface could not be measured. Instead, the measurements point to a thicker slip zone. The entire clay mass was significantly disturbed, which would indicate a zone failure. The amount of yield-induced excess pore pressure in the clay was significant. The time-dependency of the pore pressure and deformation response was also very apparent, as the excess pore pressure and displacements increased at an accelerating rate prior to failure, at a steady external load. The best indicators for an impending embankment failure seem to have been settlement under the embankment and lateral displacements in the soil. For monitoring, these should be supplemented with pore pressure measurements.