The purposes of this study were to clarify the structure of subjective information regarding movement of sprint running and to compare these structures among four groups: sprinter, distance runner, jumper and thrower. Thirteen hundred and six college athletes participated in this study. The first questionnaire was designed to determine the key items to investigate the structure of subjective information regarding the movement of sprint running. Through the administration of first questionnaire, 224 items were chosen. The degree of importance of each item was assessed on 7-point scale. MDSCAL was applied to dissimilarity matrix by each part of the body: head, arms, torso and legs, separately and the extracted dimensions were rotated orthogonally based upon Normal Varimax cirterion idea. Moreover orthogonal Procrustes rotation and Schonemann's general Procrustes rotation were applied in order to facilitate comparison and similarity of con-figurations and dimensions of these different groups were evaluated by coefficient of congruence, root mean square deviation and sum of squared error. The main results were as follows: 1) Three-dimensional spaces for head and torso, and 4.dimensional spaces for arms and legs were constructed. Stressl's were from 0.141 to 0.203 and explanatory rate were from 0.602 to 0.816. 2) The Euclidean distance model was more appropriate for representation of the structure of these information than factor (vector) model because of efficiency of explanation and interpretability. 3) Strong similarity between most of dimensions for different groups was observed in all parts of the body, so it could be inferred that these dimensions are some of the most general