To in-vestigate the effect of basic changes in the postural determinants of the perceived upright, the subject, seated in a fully enclosed room, was rotated about a circular path, and required to adjust a rod on the front wall to the true vertical and horizontal. A total of 258 subjects were tested with and without a visual field, at two speeds of rotation. With an upright field, a shift in the force acting on the body had a very small effect on the perceived upright. At the lower speed of rotation (with the force on the body shifted by 20.5[degree]) the error in adjusting the rod to the true vertical and horizontal was only 3.1[degree][long dash]that is, the rod was off by that amt. from the axes of the field. At the higher speed (with the force on the body shifted by 33.4[degree]), the error in adjusting the rod was only 6.3[degree]. Thus, under conditions of conflict between visual and postural determinants of the perceived upright, the former is more important. In the absence of a visual field, shifting the force on the body through rotation had a much greater effect on the perceived upright. At the lower speed the mean error in adjusting the rod increased to 10.7[degree], and at the higher speed to 24.7[degree]. With a visual field present, some subjects perceived the field as fully erect, and simply aligned the rod with its main axes. For other subjects, the force acting on the body had a much stronger effect: actually causing the objectively upright field to appear tilted, and leading to adjustments of the rod which more nearly conformed with the direction of this force.