Results day 2026 has come and gone. A-Level results landed on Thursday 13 August 2026 and GCSE results a week later on Thursday 20 August 2026. Ofqual, JCQ and each of the exam boards have published the headline data and the per-subject grade boundaries.
This piece walks through what the 2026 boundaries looked like, why they move year to year, and how to read the numbers without over-interpreting them. If you're revising for 2027 and want a sense of what a grade requires, the honest answer is: something close to the last few years, give or take a few marks in either direction on any given paper. The rest of this piece explains why that hedge matters.
Grade boundaries are set fresh after every exam series. If you see "you need X% for grade 9" or "X% for A*" as a fixed rule, treat it with caution. Boundaries typically sit in a familiar range but genuinely move each year. Always check the actual boundary document for the series and paper you care about.
A-Level 2026: The headline numbers
Ofqual published its A-Level 2026 summary on 13 August. Across England:
- 28.3% of A-Level entries were graded A* or A (up from 28.2% in 2025)
- 9.6% of entries were graded A* alone (up from 9.4% in 2025) (Schools Week analysis)
- 835,950 A-Levels were awarded to 320,505 students
- Entries were up 2.7% year on year, broadly tracking a 3.7% rise in the 18-year-old population
Mathematics was the biggest single-subject driver. 42.1% of Maths A-Level entries were graded A* or A, up from 41.6% the year before. Because Maths is one of the highest-volume subjects, small moves at the top end pull the overall A*/A share upward more than a similar move in a smaller subject would.
Regionally, the top-grade gap is real. London had 32.7% of entries at A* or A. The East Midlands and North East sat at the bottom on 23.1%. That's a gap of almost ten percentage points, wider than before the pandemic (7.3 points in 2019) (Schools Week analysis). Our A-Level results day 2026 dates guide has more on how the day worked and how to interpret the summary if you're going through it retrospectively.
One important framing note from Ofqual: the standard needed for each grade was the same as in previous years, and the small rise in maths top grades reflected a slightly stronger cohort (Ofqual blog, 13 August 2026). So the small tick-ups in top grades aren't a signal that the exams got easier or that grade inflation returned. They reflect who sat the papers and how the cohort performed against a fixed standard.
GCSE 2026: The headline numbers
GCSE results published on 20 August 2026. Across the full entry in England (all ages) (Ofqual GCSE 2026 blog):
- 5.7 million results were issued across 40+ subjects
- 21.8% of entries were graded 7 or above (up marginally from 21.7% in 2025)
- 67.0% of entries were graded 4 or above (down marginally from 67.1% in 2025)
For the 16-year-old England cohort (which is the group most parents will care about, since it's the standard KS4 leaver group):
- 23.1% of entries were graded 7 or above
- 70.7% of entries were graded 4 or above
The 16-year-old figures are always higher than the all-age headline because older entries (17+, 18+) include a lot of retakes, which drag the average down.
Subject entries shifted more than grades did: GCSE Statistics was up 11% year on year, Business was up 8%, Computing was down 7.2%, and French was down 5.5%. Big single-year swings in subject entries usually reflect a mix of curriculum-mix decisions at school level and student demand.
Regionally, London led on grade 4+ (71.7%) and grade 7+ (28.6%). The North East had the lowest grade 7+ share at 17.7%, and the West Midlands had the lowest grade 4+ share at 62.5%. Same broad pattern as A-Level: London ahead, North East and Midlands trailing.
Our GCSE results day 2026 dates piece walks through the day itself if you're catching up after the fact.
How grade boundaries work in practice
This is the bit that gets misunderstood most often. A grade boundary isn't a fixed percentage. It's a raw mark, set fresh after each exam series, that separates one grade from the next.
The process in outline: students sit a paper. It's marked. The chief examiner and awarding committee look at the raw mark distribution against expected outcomes for the cohort (this is where "comparable outcomes" comes in: the regulator's mechanism for keeping standards steady year to year). They set the key boundaries (grades 7, 4 and 1 at GCSE; A and E at A-Level), and the other grades follow from those, at points that produce the intended outcome for that cohort against that paper's difficulty (Ofqual on setting A). If the paper was harder than usual, boundaries drop. If it was easier, boundaries rise.
That's why you shouldn't quote a fixed percentage. What you can say is that boundaries typically sit in a familiar range for a given subject and tier, and that if a paper felt unusually hard the boundary is likely to reflect that.
A worked example: AQA GCSE English Language 2026 was marked out of 160. The grade 9 boundary sat at 122 (roughly 76% of the total). The grade 4 boundary sat at 75 (roughly 47%). Compare that to AQA GCSE Geography 2026, marked out of 252. Grade 9 was at 204 (roughly 81%). Grade 4 was at 112 (roughly 44%).
The percentages differ because the papers differ, but neither is the "real" answer. Both are AQA-set boundaries for those specific papers in that specific series. Next year those numbers will move.
A general rule of thumb, hedged appropriately: a grade 9 in GCSE typically sits around 75-85% of the total marks on a given paper, and a grade 4 often sits somewhere around 35-55%, depending heavily on the paper. That's a rule of thumb, not a rule.
Ready to boost your grades?
Join 1M+ students who have used Cognito to ace their exams.
Get started for free!
Where to find the actual boundary documents
Every board publishes its own grade boundary documents. If you want the real number for a specific paper:
- AQA: aqa.org.uk grade boundaries page
- Pearson Edexcel: qualifications.pearson.com grade boundaries downloads
- OCR: ocr.org.uk grade boundaries page
- WJEC / CCEA / Qualifications Scotland: each board's own boundaries page
JCQ (the joint council) publishes cross-board interactive dashboards, which are useful for looking at national headline results. But for a specific paper's boundary you want the board that set the paper.
For 2025 series comparisons, we have breakdowns for GCSE Edexcel 2025 boundaries, GCSE OCR 2025 boundaries and A-Level Edexcel 2025 boundaries. These are worth looking at side by side with the 2026 numbers to see how much movement there is in any given subject.
Why boundaries move year to year
Three main reasons.
Paper difficulty. The dominant factor. If a paper is harder than usual, the boundary drops. If it's easier, the boundary rises. This is the mechanism that lets Ofqual honestly claim that a grade means the same thing year to year even when the numeric threshold moves.
Cohort effects. A larger or stronger cohort will produce more high marks against the same paper. Ofqual controls for this in the outcome-setting process, so the top-grade share stays roughly stable, but individual paper boundaries still reflect where the raw marks fell.
Comparable outcomes. This is the regulator's mechanism for keeping standards stable across series. It uses prior attainment data (typically KS2 results for GCSE, or GCSE results for A-Level) to project what the current cohort's outcomes should look like against a fixed standard. Boundaries are then set to hit that projection.
The upshot: don't chase specific percentages when you're revising. Aim to secure the content and technique. If you're routinely getting 80%+ on past papers under timed conditions, you're in strong shape for a top grade in most subjects, whatever the boundary lands at.
Scotland: A different calendar
Quick note if you're in Scotland or you're following Scottish results. Qualifications Scotland (which replaced the SQA) operates on its own calendar. National 5s, Highers and Advanced Highers were released on Tuesday 4 August 2026, over a week ahead of the English A-Level day. Qualifications Scotland publishes its own grade boundaries via its own portal, and comparisons with English or Northern Irish results are only broadly meaningful. If you're a Scottish student, qualifications.gov.scot is the source you want.
What this means if you're revising for 2027
A few takeaways.
- Don't fixate on hitting a specific percentage. The percentage that gets you a grade 9 or an A* in 2027 will not be identical to 2026. Aim for mastery of the content.
- Past paper practice is the closest thing to a real signal. Use the mark schemes and examiner reports to see where marks are awarded.
- Boundaries are published on results day (AQA grade boundaries). If you're comparing your performance to a boundary, use the published document for that series and paper.
- If your school flags a subject-level specification revision, work to the current spec rather than an older revision guide.
Cognito's A-Level Biology course notes and GCSE Maths course notes map to the current specifications for those boards. If you're picking a revision resource, that's the version-check to run first.
