This article examines the relation between the perception of one's own body position and the distribution of fluid along the subject's spinal (z-) axis. Two experiments are reported. The first one is a replication of the Vaitl et al. study [J. Psychophysiol. 27 (1997) 99] which has shown that changes in shifts of blood volume into or out of the thoracic cavity induced by lower body positive pressure (LBPP: +30 mmHg) or lower body negative pressure (LBNP: -30 mmHg) exerted on the lower body led subjects to feel tilted head-up or head-down, respectively. The second experiment was designed to differentiate between the influence of the otoliths and of the changes in fluid distribution on the perception of body position by means of a sled centrifuge in combination with LBPP and LBNP. In both experiments, changes in blood distribution within the thoracic cavity were measured by impedance plethysmography. Forty healthy volunteers (17 females) participated in experiment 1. They were positioned on the side (right-ear-down head position) on a tiltable board which the subject and the experimenter could tilt via remote control around the subjects' z-axis. Subjects were asked to rotate the board until they felt they were in a horizontal posture. The results clearly show that the perception of posture is influenced by the shift in blood distribution. During LBNP subjects perceived being tilted head-up, whereas LBPP led them feel tilted head-down. Thus, the results obtained in the 1997 study were replicated. Fourteen males volunteered in experiment 2. They were positioned on the sled on a centrifuge in the same manner as in experiment 1. The sled could be moved via remote control by both the subject and the experimenter. While the centrifuge rotated (omega=2 pi times 0.6 rotations per second) the subjects were asked to move the sled until they felt they were in a horizontal position. As in experiment 1, shifts in blood volume were induced by LBPP and LBNP. The distance between the binaural axis (position of the otoliths) and the centrifuge axis served as dependent measure indicating the subjective horizontal position. Due to the additional centrifugal forces exerted on the body the shifts in blood volume were more pronounced than in experiment 1 where only gravitational forces were produced. The changes in the perception of posture were influenced by both the otoliths and the fluid distribution in such a way that both interact in a compensatory manner. These results again corroborate the evidence that afferent inputs from the cardiovascular system play a major role in the perception of the body posture. This phenomenon of graviception needs to be further elucidated with respect to the origins of the afferent inputs and the site and type of graviceptors (mechanoreceptors) involved.