The purpose of this study is to examine effects of the admission system to higher education on background diversity of students and study efficiency. By using data from a longitudinal project, the students admitted to medical education on different admission grounds are compared. The results indicate that admissions based on upper secondary grades best promote the goal of diversity. In addition, those students perform nearly as well as those admitted via step-wise procedures. The step-wise procedures promote study efficiency, but tend to favour applicants with upper-middle-class and Swedish backgrounds compared to those admitted via grades. Those admitted on the basis of SweSAT scores perform more poorly than those admitted via grades and step-wise procedures. Furthermore, the results indicate that admission via SweSAT scores contradict rather than promote the goals of increasing social, educational, and national diversity in higher education. © 2006 Scandinavian Journal of Educational Research.