An actuarial model applied to the selection process of a special medical school program at the University of Maryland School of Medicine was tested. The 77 students in the study sample were admitted to the university's Fifth Pathway Program, which is designed for U.S. citizens who completed their medical school training, except for internship and social service years, at a foreign medical school. This program involves a year of supervised clinical training in university affiliate hospitals to prepare students to enter a residency training program. Application data were subjected to multiple regression analysis using post-program clinical performance ratings and test scores as the criterion variables. Preliminary formulae for selection have been developed and several application variables have been combined to predict future performance. The variables of most interest are the Medical College Admission Test scores, which have been shown in previous studies to predict basic science but not clinical performance in medical school. Implications for further research, including a cross-validation study, are discussed. (Author/SW)