The use of dynamic driving simulators by researchers and engineers is expanding rapidly due to the many advantages of simulation as compared to traditional methods of investigation, and also as a result of rapid technological progress over the past decades. However, because every simulator is a prototype and there are no set manufacturing standards, their validity and possibilities for development can be questioned. This paper reviews recent developments in the design of simulators for use in research, and describes the reasons and means supporting these innovations. After defining the need for multisensory stimulation, we describe the main sensory interfaces (visual, motion, tactilo-kinesthetic, and sound stimulation), while paying particular attention to the problem of longitudinal movement cueing and its consequences on braking behaviour. We also examine the limitations of simulators in relation to simulator sickness. This review demonstrates the need for in-depth research, not only to assess the validity of simulators and their technological changes, but also to guide their future development. A chart summarizes the features of current driving simulators and describes a wide variety of configurations.