Many traffic accidents are due to winter conditions like slippery roads and limited visibility. The road administrators put a lot of effort into snow removal and de-icing the roads, and the vehicle manufacturers have been working with functionality to support drivers in winter conditions for decades. Many issues of driving in winter conditions originate in drivers’ behaviours such as risk taking and lack of awareness. Studying drivers’ behaviour in winter conditions in general, and the effect of various countermeasures of the vehicle, would increase the understanding of the underlying mechanisms and could possibly be used to reduce the accident risks. Motion-base driving simulators are tools frequently used for driving behaviour research. However, the validity of the results of such studies depends to a large extent on the realism of the simulation. The purpose of this project was to study winter simulation with the aim to improve the realism of simulator driving in winter conditions. Driving in winter is in many ways different from driving in summer. The difference can to a considerable extent be explained by the tire-to-road interaction. Winter driving is typically characterized by softer motion and slower development of tire forces. In the present project two aspects of the motion have been studied, the motion feedback in the simulator and models for tire-to-snow behaviour. Vehicle motion during winter driving is characterized by large vehicle body slip angles and the associated yaw motions. Thus, understanding the importance of yaw motion feedback is essential to simulate winter conditions. A study was designed to investigate the impact that lack of yaw motion will have on the driver, and if it alters his driving behaviour. It was found that the yaw motion alters the driving behaviour, implying that it contains vital information for a wide range of driving situations. A second study was carried out to investigate if the rotation centre, which is a consequence of the yaw and lateral motions, can be used to present valuable information about the vehicle state to the driver. Indicative results suggest that the rotation centre of the motion is valuable to the driver, but further studies are needed. The softer and slower motions of winter driving suggest that pre-positioning of the simulator’s motion platform could be performed to improve the motion envelope. A third implementation study performed in the project suggests this.