This paper re-evaluates traditional approaches to the study of children's drawings in the light of current process-oriented research. A tentative synthesis of accounts of developmental stages in drawing is followed by an appraisal of drawings analysis in applied psychology. Main features of the new approach are outlined. (Author)