The observed decline in academic performance among 15-year-old students, as reported by recent OECD-PISA surveys, alongside the democratisation of university admissions, points to a potential decrease in the marginal academic proficiency of incoming higher education students. Paradoxically, grades at the tertiary level have either remained stable or risen, suggesting the presence of grade inflation, i.e. a relaxation in grading standards. This trend may have been amplified by the 'tolerance' or 'grade leniency' requested during the COVID-19 pandemic. This paper explores three key areas: first, the existence and degree of grade inflation from 2012 to 2022; second, whether the pandemic exacerbated this inflation; and third, the effect of lenient grading on student outcomes, including pass rates, exam participation, graduation likelihood and dropout rates. Drawing on data from 28,520 students across 2000 courses at a major European university with an open admission policy, panel data fixed effects models reveal an average modest trend of grade of inflation (around 0.012-0.024 standard deviations), which significantly accelerated during the pandemic years, with increases reaching nearly 10 times the usual rate. A Coarsened Exact Matching analysis, used to account for shifts in student demographics and abilities over time, shows significant differences between weak (i.e. individuals with an average number of retake exams above 1.5 or bottom 40% of the grade distribution) and strong students (i.e. average number of retake exams below 1.2 or top 35% of the grade distribution). Grade inflation primarily benefited weaker students, while stronger ones remain unaffected.