The main purpose of the TRAINER project was to develop a new cost-effective Pan-European driver training methodology, based on computer based interactive multimedia and simulator technology, which paid significant attention not only to gain experience of driving and handling the vehicle, but also to the enhancement of risk awareness of learners drivers. For this purpose simulation tools were developed. A number of scenarios for application in the simulators were implemented. The scenarios addresses the most important needs of learner drivers, based on accident statistics and an extensive literature review. They have been structured in accordance with the four hierarchical levels of the GDE-matrix. In order to be able to assess the TRAINER tools impact on traffic safety, driver behaviour was measured in an experiment in the VTI research simulator. The drivers ' behaviour is , however, not only influenced by their driving performance skills, but also on their hazard perceptual abilities. This chapter covers one of two experimental studies carried out at VTI with the purpose of evaluating the effects of the TRAINER simulators on traffic safety. The experiment presented here is measuring driver behaviour and was carried out in the VTI research simulator. The other experiment was carried out in real traffic, where hazard perceptual skills were assessed by data from eye-tracker recording of visual search strategies of the driver. The results show significantly safer driver behaviour in the trial group when driving in fog and when responding to a SMS on a mobile phone. There were no significant differences in scenarios where a bus or where a moose unexpectedly appeared on the road. The results show that it is possible to improve driving behaviour by including simulator training in driver education. More research and development is, however, needed in order to design scenarios for better safety effect in more situations.