Medical schools select applicants; applicants who hold two or more offers also select medical schools. In this study we examine a large cohort of applicants applying for admission to UK medical schools in October 1991, and show that when presented with a choice, applicants are more likely to select an interviewing school over a non-interviewing school. Preferences for individual schools are examined in detail using the Bradley-Terry method for paired comparisons. The analysis shows that the preference for interviewing schools cannot be explained in terms of a preference for Oxford and Cambridge (which both interview), nor is it due to confounding by more non-interviewing schools being in the north of the UK. Nevertheless, interviewing status does not account for all the preferences expressed by applicants. The preference for interviewing schools is perhaps explained by applicants feeling closer and more involved with schools which they have visited, which have talked to them personally, and which have selected them individually, rather than on the basis of an impersonal application form.