The lack of tested models of the instructional process is the prime hindrance to significant management science in educational research. The nature of education, with long timelags between ultimate performance and present activity, makes the modeling most difficult. This paper reports on one of several efforts which have been made to develop quantitative models of the educational process. The model is developed for future inclusion into a school district budgeting and program-planning model. The history of the model's development and some of the problems encountered are discussed. Practical implications for further management science in educational research are also drawn. (Author/HW)