A review of the literature on automated psychological testing shows that most research and development in this area is based on a one-test, one-psychologist model. In this model, the functions of test administration, scoring, and interpretation are thought out in terms of specific tests presented on an individualized basis. However, more complete and sophisticated psychological assessment can take place with a multi-test, multi-team model. This model makes extensive use of time-sharing, interactive, terminal-oriented computer systems. Methodological investigations and research and development work on automated testing in the last decade are reviewed in terms of the three dimensions of testing: administration, scoring, and interpretation. A computerized information management system for the storage and retrieval of student evaluation files appears to be necessary. Such a system would allow varying and appropriate reports to be generated for psychologists, teachers, counselors, etc. The capability of computers to analyze and accept natural language input must be further developed before testing can become fully automated. (JK)