The present state of research in the area of automated teaching and the application of automated teaching devices were reviewed in terms of issues relating to programing, to machine variables, and to studies comparing conventional instruction with automated instruction. Some basic issues were singled out for consideration. Such problems as prompting, confirmation, reinforcement, size-of-step, and branching were considered. The design and construction of machines were judged to be a secondary issue; the question of programing takes precedence. The problem of measuring the effects of automated teaching was considered to be crucial, since it is bound to the question of what is to be learned. It was also pointed out that job analyses of learning tasks must be done and measurement problems solved before valid comparisons between automated and conventional instruction can be made. (Author)