A technique is described for private liberal arts colleges to assess the probable matriculation behavior of admitted applicants. By performing a multiple discriminate analysis of data obtained from applicants admitted during a previous admissions season, and applying the information to a decision-making strategy used in political elections the procedure can be used to increase the institution's yield. The method calls for complete and detailed data on applicants from the immediately preceding admissions season which provides the basis for differentiating between those who enrolled at the target institution, and those who matriculated at competitor schools. In addition, information from the analysis of this data set must be applied to findings on currently admitted applicants in order to build a prediction function consisting of the factor weights produced from the initial data with the available data and currently admitted applicants. The information, subject to the prediction equation, produces categories of candidates with a high, intermediate, and low probability of matriculating at the school. Included in the paper is a detailed description of the application of this system to the admissions process at Whitman College (Washington) in 1988. (JB)