The amount and rate of pronation and supination have been the subject of interest to runners for some time. Exact determination of the motions has been hampered by their complexities and use of a two-dimensional data collection protocol. "Rearfoot motion," measured by determining the projection of the angle between a line on the posterior of the shank and a line on the heel, has been a common approach. This projection measures a rotation about a laboratory axis and not a body segment axis and has a potential of error due to projection onto a plane. The angle measured in rearfoot motion is not the true angle between these lines in space and has projection distortion errors which are compounded during plantar and dorsiflexion and medial and lateral foot rotations. The rearfoot motion angle, however, does approximate foot inversion-eversion during much of stance phase. The proposed change in research protocol allows analysis of the three-dimensional position data of targets to construct a "joint coordinate system" which gives more accurate data on inversion-eversion and data on plantar-dorsiflexion and medial-lateral rotation of the foot. This analysis may allow the examination of measurable differences between individuals and running shoes of various design.