Background: Haptic guidance is a continuous form of haptic feedback where the driver obtains information through small corrective forces on control inputs, such as the steering wheel. Haptic guidance for automotive steering has shown promising results such as improved primary task performance, reduced visual workload and reduced control activity. Despite these promising results, it remains unknown how much information on curve negotiation is provided by haptic guidance when combined with full visual or reduced visual information. Objective: The goal of this research is to gather empirical data to investigate the capability of haptic guidance to present essential steering control information for curve negotiation to drivers when essential visual information is removed. Method: Previous research identified a far and near visual point for providing drivers with essential visual control information. This research removed essential visual control information while haptic guidance was presented to the subjects. The visual information was removed by occluding all visuals above 7.5o down from the true horizon for the near visual condition and all visuals beneath 1.68 down from the true horizon for the far visual condition. This corresponds to 0.34 seconds and 1.53 seconds look-ahead time respectively when driving at 20 ms-1 . It was hypothesized that the designed haptic guidance would be able to compensate for this loss of visual information. Results: While using haptic guidance, driving performance and control activity show a significant performance increase for all visual conditions. For the experimental conditions tested, control effort only increases for full visual and slightly for the far visual condition. The summed absolute lateral position is increased to better than baseline performance for all visual conditions, however time to line crossing performance is not increased to better than baseline performance for the near visual condition. Conclusion: Based on the empirical data it can be concluded that haptic guidance is capable of providing drivers with essential control information, although driving performance does not match that of full visual feedback without haptic guidance. It also appears that drivers are more likely to accept haptic guidance if essential visual information is removed. Application: With this empirical data new mathematical and cognitive driver models can be developed that incorporate haptic feedback which can be useful in the future development of autonomous systems based on human driving behavior.