The ability of the Perceptual-Motor-Rating Scale to indicate changes in regular classroom performance resulting from individualized perceptual-motor training by physical education specialists was statistically analyzed. Efforts were made to identify factors, confirm their independence, optimize scoring procedures and assess reliability. Preliminary statistical analyses of pre-post test results yielded four factors; and for 344 first graders, several significant differences between project and control schools, and among classes. Additional analyses of results for 601 children (grades one through three) confirmed the emergence of these four independent factors when full loadings on all items were used, regardless of magnitude; the independence of factors vanished, however, with two attempts at single-digit item loadings--perhaps because several items loaded heavily on more than one factor. Further investigation of the reliability and validity for project evaluation of a global score has been planned. This should make quantification of observable changes in children's behavior less computer-dependent and, thus, the instrument should become more usable. The scale itself and a history of the program for which it was developed, the Perceptual-Motor Development Project, are appended. (Author/CP)