The use of subjective probability as a theoretical model for enrollment forecasting is proposed, and the results of an application of subjective probability to enrollment forecasting at the University of Toledo are reported. Subjective probability can be used as an enrollment forecasting technique for both headcount and full-time equivalent using the following steps: (1) assume the various subpopulations are normally distributed; (2) determine the mean and standard deviations of these subpopulations; (3) sum the means and variances of the subpopulations to obtain the mean and variance of the university forecast; and (4) perform a linear transformation on the headcount distribution to obtain the full-time equivalent distribution. Major advantages of this method are: it allows the study of homogeneous subpopulations; it allows comparisons with other methods, since the variability is known; and it allows probability statements regarding various headcount and full-time equivalent outcomes. Application of the subjective probability method to enrollment forecasting at the University of Toledo resulted in a difference of 0.4 percent between the actual and the forecasted headcount enrollment and a difference of 0.3 percent between actual and forecasted full-time equivalent enrollment. In comparison with the conventional forecasting techniques also employed in the test study, the forecasting errors using the subjective probability were the smallest. (SC)