The COVID-19 pandemic has affected the way we learn, teach and support students in many different ways. In particular, the impact of the pandemic on the mathematical capabilities of students arriving at university is now starting to be seen and reported. Students arriving at UK universities in 2021, and having completed their A levels in 2021, had received disruption to almost all of their A level learning and it seemed likely that this might cause them to find the transition into higher education more difficult. Despite the disruption, the number of students achieving the top A level grade has more than doubled since 2019. In this paper, the data from a diagnostic test that has remained the same for the past 30 years are considered to measure the mathematical preparedness of students starting their courses at one UK university in September 2021, updating the work of Hodds (2021). Furthermore, an ordinal logistic regression model predicting A level grade based on test performance and other factors is developed to help determine if grades were perhaps inflated as suggested. The results show that students entering in 2021 are significantly less prepared for the mathematics on their courses than their colleagues who entered in 2020 and pre-pandemic. The regression model predicts a higher probability of being awarded an A[superscript *]/A for students with A levels awarded in the pandemic, despite far lower scores on the diagnostic test, and in particular in 2021 when A levels were awarded through teacher assessment of the topics covered and not necessarily the whole A level syllabus.