One potential approach to the authentic assessment of what students know and can do in science is concept mapping. A concept map is a graph consisting of nodes representing concepts and labeled lines denoting the relation between a pair of nodes (concepts). The external concept map constructed by the student is interpreted as representing important aspects of the organization of concepts in that student's memory (cognitive structure). This review of the literature found that concept mapping techniques varied widely, with no less than 128 possible variations identified. Methods for scoring maps varied almost as widely. The review leads to the conclusion that an integrative, working cognitive theory is needed to begin to limit this variation for assessment purposes and to serve as a basis for psychometric studies of the reliability (generalizability) and construct validity of concept maps. Other issues arising from large-scale use of mapping techniques are explored, including the importance of students' familiarity with and skills in using concept maps and the possible negative impact of teachers teaching to the assessment, thus requiring students to memorize concept maps provided by textbooks or themselves. Ten figures and two tables present review findings. (Contains 41 references.) (Author/SLD)