A vibrotactile display, consisting ofeight vibrating elements or tactors mounted in a driver's seat, was tested in a driving simulator. Participants drove with visual, tactile and multimodal navigation displays through a built-up area. Workload and the reaction time to navigation messages were measured for normal and high workload conditions. The results demonstrated that the tactile navigation display reduces the driver's workload compared to the visual display, particularly in the high workload group. The fastest reaction was found with the multimodal display. It was concluded that this study quantitatively supports the claims that a localised vibration or tap is an intuitive way to present direction information, and that employing the tactile channel may release other heavily loaded sensory channels, therefore potentially providing a major safety enhancement.