The goal of this study was to determine whether vestibular information plays a crucial role in determining how the CNS maintains posture and controls movement. Resting posture and locomotion were compared in wild type and vestibularly deficient mice (IsK -/- mutant) using X-ray and high-speed cineradiography (250 frame/sec), respectively. In IsK -/- mutant the semicircular canals cristae as well as the macula of the utriculus and sacculus exhibited significant signs of degeneration. In contrast, detailed analysis of the vestibular nerve fibers and Scarpa ganglion cells indicated that there was no degeneration in these structures as compared to wild type mice. This lack of afferent degeneration strongly suggested that central vestibular network remained intact in IsK -/- mice, consistent with previous lack of evidence for IsK transcripts in brain. Standard behavioral assays and vestibulo-ocular reflex testing were used to verify the complete absence of vestibular function in IsK -/- mice. The principal findings were that: 1) the S-shaped resting posture of the vertebral column in the sagittal plane was characterized by the fact that both wild type and mutant strains positioned i) their heads near the extreme point of flexion in the atlanto-occipital articulation, ii) the cervico-thoracic junction at its extreme point of flexion and iii) the cervical column upright, thereby reducing the available degrees of freedom; 2) locomotion in wild type mice was characterized by i) the linear progression of limb movements which followed the lateral sequence and ii) the stability of the head and entire extended column relative to space; 3) in contrast, locomotion in IsK -/- mice was characterized by episodes of circling during which the forepart of the body maintained an S-shaped rest-like configuration. Hence, vestibular information was mandatory for providing a reference for head position in the sagittal plane during locomotion. In contrast, vestibular information was not required for development of the stereotyped configuration, which characterizes resting posture. Our findings support the hypothesis that mice, and presumably other quadruped vertebrates, adopt a finite number of skeletal configurations, which limit the number of degrees of freedom needed to control movement. Furthermore, we propose that vestibular inputs provide feedback required to signal the completion of a planned trajectory. When Isk -/- mice intend to change their heading direction this information is lacking, and therefore they continue to rotate.