In considering and evaluating approaches to the admission of college students, the usual approach is to try to measure past academic achievement and primarily verbal and math ability on the assumption that these abilities will predict subsequent college academic grades and achievement. These measures do predict classroom achievement, though far from perfectly so. It is also the case that most universities claim to develop students in areas not well represented by classroom grades such as leadership, social responsibility, integrity, multicultural appreciation, and others. In our work, we have adopted a model employed by industrial/organizational psychologists in personnel selection. We began with a "job analysis" of the "job" of undergraduate students. We developed a list of expectations universities claim to have of students and derived a list of constructs that were hypothesized to be essential to success. This set of constructs has been central to the development of a series of measures we use to assess student potential as well as a set of outcome measures that we believe is a better representation of the totality of relevant college student outcomes. (Contains 5 tables.)