The 2019/20 Level 4 mathematics cohort at the Nottingham Trent University sat a full set of mid-year assessments in January 2020 under completely normal circumstances. However, the COVID-19 lockdown meant that their end of year assessments, along with all of their teaching and learning from March 2020 onwards, moved fully online. This has given us a unique opportunity to understand how the same cohort perform in contrasting situations. In this study we consider the issue of attainment gaps and find that the attainment gap closed in this cohort for black and minority ethnic students but that students from a lower socio-economic background may have been put at a disadvantage by the move to online teaching, learning and assessment. We use a linear mixed effect models approach to present statistical evidence to support these two claims as well as investigating the specific aspects of the move online, which may have caused these results.