In this study an attempt was made to obtain a developmental dimension for estimating longitudinal development on the basis of cross-sectional data. To check the validity of this approach, the cross-sectional data were compared with true longitudinal data. Forty-three public school children were given a series of conservation tasks in four assessment periods, each 15 months apart, during their first four years of school. Results indicated that: (1) cross-sectional data on cognitive development, when scalogram analyzed, can predict longitudinal direction of development; (2) such developmental scales may better characterize a child's present level of development than deviance scores on IQ-tests; (3) such scales are useful for studying the development of individuals as opposed to groups; and (4) these scales may be helpful in comparing individual rates of development. (ED)