Despite considerable attention, the process of attempting to select the finest applicants available from a pool larger than the number of students that can be admit-tcd still presents many problems to the admissions committces. Perhaps the hardest of these problems to deal with is the fact that it is not clear what criteria definc the finest applicants because it is difficult to measure the quality of a medical school graduate in terms of his ability as a physician (1). Therefore, in practice, most ad-missions committees do not accumulate cvidence that can be used to view current admissions records for predictive correlation with this principal objective of medi-cal education. Instead it has been standard practice for admissions committees to substitute criteria such as academic performance in medical school (rank in class by a combination of test scores) when attempt-ing to assess the predictive values of the several kinds of credentials commonly evaluated. Given this situation, it is perhaps not surprising that, in face of a be-