The problem of improving the quality of incoming college students and practicing fairness in admissions was studied at Boston College. The study sample was over 2000 accepted applicants who completed an admissions questionnaire. A systematic admissions procedure is described that provides an index of applicant "quality," and the rating process is explored using a path model to measure the contributions of background and achieved characteristics of applicants to their rating. The way questions of bias may be raised and pursued is considered, and applicants are profiled to show how the effects of policy adjustments may be monitored. Six segments were defined by crossing three levels of quality (top, high, or medium) with two values of enrollment status (matriculation or nonmatriculation). Factor analysis was used to model the differing images of Boston College across the six segments. Discriminant analysis was applied to represent two dimensions, "decision" and "quality" in ratings of both Boston College and of the named competitor. Some specifics of a differentiated policy are considered as examples, and implications of this integrated approach are discussed. (Author/SW)