The role of the vestibular system in acquisition and performance of a spatial navigation task was examined in rats. Male Long-Evans rats received sham or bilateral sodium arsanilate-induced vestibular lesions. After postoperative recovery, under partial water-deprivation, rats were trained (16 trials/day) to find a water reward in one corner of a black square enclosure. A cue card fixed to one wall of the enclosure served as a stable landmark cue. The orientation of the rat at the start of each trial was pseudo-randomized such that the task could not be solved by an egocentric response strategy. Rats with vestibular lesions acquired the task in fewer trials than the sham treated control rats. Vestibular lesions did not influence the motivation or motor function necessary to perform the task. Performance of sham rats was maintained during probe trials in which the cue card was removed from the enclosure, while lesioned rats were markedly impaired. Rotation of the cue card (+/-90 degrees ) caused an equivalent shift in corner choice behavior of the lesioned rats. However, sham rats often disregarded the rotated cue card and made place responses. These results suggest that the vestibular lesions disrupt idiothetic navigation or path integration and render navigational behavior critically dependent upon external landmarks. These results are consistent with the navigational abilities of humans with bilateral vestibular dysfunction.