Evacuation of trains in tunnels is currently taking place in diverse ways. One of the methods implies that passengers leave the train along the railways on elevated walkways. The knowledge about the impact of elevated walkways on the safety level is today very limited. As more and more elevated walkways are designed in rail tunnels, it is important that studies and evacuation trials are made to investigate how those affect the safety level of the evacuation. This master thesis’ project aims to investigate the relationship between low light conditions, people's behaviour and ability to evacuate a train on an elevated walkway. To investigate this, the core in the work consisted a practical evacuation trial which purpose was to study people’s movement on a raised walkway under different illumination levels: 200 lux, 5 lux and 1 lux. A total of 16 escape trials were performed as controlled evacuations on a simulated elevated walkway with the measures 1.2x20 meters with a level difference down to the ground plane of 1.24 meters. The result of the evacuation trial showed that the intensity of illumination played a significant role in both the movement speed, the person flow and the peoples distance to the edge. The peoples flow and speed did generally decrease during the partial trials performed during the weaker light intensities, 5 lux and 1 lux, compared to partial trials performed during 200 lux. A general result regarding the effect of light intensity on the people’s distance to the edge is that during the partial trials performed with the weaker light intensities, 5 lux and 1 lux, more people chose to go further from the edge. The conclusion of these results is that a minimum brightness in tunnels should be 1 lux, but a stronger illumination should be sought to increase the safety of the passengers in case of evacuation. Regarding the learning effect on the trial procedure, it can be seen from the results that the more trials carried out, the closer the edge the people went combined with an increased speed and flow. The people became comfortable in the environment and hesitated less, which generated a source of error in the result. A conclusion of the practical evacuation trial is that by conducting a trial in this type of environment, we were given the opportunity to study the problems as well as the complexity that an evacuation may imply. The experiment further provided valuable information and knowledge about the problems that may arise in an evacuation, both from a technical and behavioural perspective.