Mathematics Support Centres are to be found in various forms in the majority of UK higher education institutions. They have been established in order to ease widespread and serious difficulties that a significant number of students have with mathematics, particularly at the school–university transition. They usually offer mathematics and/or statistics support to students across the full range of disciplines studied. Anecdotal evidence suggests that those students who make good use of such centres are not just those who struggle with mathematics. Many frequent users are quite competent and simply want to do better. The study reported here describes and analyses data from one cohort of engineering students. A novel aspect is the quantification of the proportion of support centre visitors who fall into these, and other, categories. We conclude of the cohort in the study, mathematics support has improved the pass rate by ∼3%. Of the failures, about half (∼4% of the sample total) could well have passed had they attended the mathematics support centre regularly. Furthermore, the majority of those attending were not students who were in danger of failing. This has important implications not only for the design of mathematics support provision, but also for the performance of the high fliers. The methodology offers one way tackling the difficult task of evaluating the effectiveness of mathematics support initiatives.