The differential effects of vision on motion sickness in cars were tested under real road conditions using linear accelerations, in order to confirm earlier laboratory results on visual modulation of vestibular nausea induced by angular accelerations of the body. The 18 voluntary subjects were exposed to repetitive braking maneuvers (linear accelerations: 0.1-1.2 g) on a highway. The simultaneous visual stimulus conditions for the 3 separate days were: I) eyes open, visual control of car motion; II) eyes closed; III) eyes open, artificial stationary visual field (reading). The severity of motion sickness (magnitude estimation 1-10) was a function of the visual stimulus condition with significant differences among these conditions: I) moderate nausea (less than 1) with adequate visual motion perception; II) medium nausea (approximately equal to 2) with eyes closed and somatosensory-vestibular excitation only; III) strong nausea (greater than 5) with conflicting sensory input, when vestibular acceleration is in disagreement with the visual information of no movement. Providing ample peripheral vision of the relatively moving surround is the best strategy to alleviate car sickness.