Abstract
The implications of recent studies of the selection process for entry to medical school are discussed. It is argued that there are two types of factors that need to be considered in selection; first, those that are essentially concerned with policy decisions in relation to entry, e.g. sex or nationality; and, second, those that are to do with individual characteristics. In relation to this second group, once minimal criteria are satisfied, random choice seems as fair as any method of selection, as a means of choosing among applicants: none appears to be sufficiently predictive of performance to be suitable as a means for selection in themselves.