Advocates of using a US-style SAT for university selection claim that it is fairer to applicants from disadvantaged backgrounds than achievement tests because it assesses potential, not achievement, and that it allows finer discrimination between top applicants than GCEs. The pros and cons of aptitude tests in principle are discussed, focusing on school-proofness claims, concluding that it is invalid to assess suitability for university using measures purportedly immune to study, and unfair to allocate opportunities according to qualities acquired by chance. Considering empirical findings, aptitude tests' claims to school-proofness and their power to predict academic achievement - relative to achievement tests - are found wanting. Methods of evaluating pupils' achievement within an educational context are discussed, including a national system for ranking university applicants that accounts for absolute achievement and educational context. Pupils' ranks would form the basis of universities' first sifts, allowing applicants from diverse socioeconomic backgrounds to compete fairly for university places. (Contains 8 endnotes.)