There is a growing demand for admissions tutors in UK universities to use more rigorous and systematic decision-making methods in recruitment processes. A key barrier to achieving this goal is the lack of a coherent and definitive body of knowledge upon which such methods can be built. The yardstick used for the majority of applicants is A2-level performance measured through the grades that are attained, often translated into a points score. However, A2-level performance is widely thought to be a crude and inaccurate way of predicting future academic performance. The aim of the research reported in this paper was to inform the practice of admissions tutors by improving the basis upon which decisions are made. Specifically, the goal was to provide admissions tutors with evidence to support their decision-making processes by providing concrete data upon which recruitment decisions can be made. The research studied the first year (Level 1) performance of one cohort of students studying mechanical engineering programmes at the University of Leeds. The focus of the research was on the relationship between A2-level performance and other information that is available to admissions tutors when they are selecting students, and Level 1 university performance. The results of the research indicate that, for applicants with A2-levels, (1) only 28% of the variance in their university performance could be attributed to their A2-level performance and (2) admissions processes should focus on the A2-level grades that are attained rather than the A2-level points scores that are frequently used. (Contains 1 note, 1 table, and 17 figures.)