The application of multiple linear regression to the problem of identifying appropriate criterion variables and predicting enrollment in college courses during a period of major rapid decline was studied. Data were gathered on course enrollments for 1972-78 at Kent State University, and five independent variables were selected to determine the most likely indicators of future enrollment trends. The variables were: total departmental majors in elementary education, elementary education majors plus all those enrolled in service courses and allied majors, the number of elementary education freshmen enrolled for each year, total enrollment of the college of education by year, and total university enrollment. Data were analyzed using Multiple Regression Analysis: Subprogram--Regression from the "Statistical Package for the Social Sciences" (SPSS; Nie, Hull, Jenkins, Steinbrennar, and Bent, 1975). Total college enrollment proved to be the best single predictor. The technique proved to be 96 to 100 percent accurate in estimating course enrollments in seven of the 10 courses. It is also a valuable means for data-based decision-making and long-range planning when faculty committees must advise on administrative decisions. (Author/SW)