Why Do Mock Grades Drop in Real Exams? The 2026-27 Answer

A mock exam grade and a real GCSE or A-level grade are not measured against the same number. AQA states it “can’t decide grade boundaries until we have seen how students have performed in the exams” — real boundaries exist only once a national cohort sits and is marked. A mock’s boundary is chosen by the school instead: one Staffordshire secondary reuses 2023’s real boundaries with a stated 5% uplift, one of four different named methods below.

Do Exam Boards Set Grade Boundaries for Mock Exams?

No. AQA’s own myth-busting page states it directly: “We can’t decide grade boundaries until we have seen how students have performed in the exams.” That statement is about the real, live paper — the one every GCSE or A-level student in the country sits on the same day. A mock exam has no equivalent national sitting, so there is no equivalent national boundary for AQA, Pearson Edexcel, OCR or any other board to publish. Ofqual’s current guide for schools and colleges, updated 3 February 2026, makes the same point from the other direction: when setting expectations “based on mock exam performance” for the reformed GCSE French, German and Spanish specifications, schools “should not rely on how students performed compared with past grade boundaries.”

This is not a new position. Cath Jadhav, Ofqual’s Associate Director of Standards and Comparability, wrote in a 2017 Ofqual blog post that “even in well-established qualifications, grade boundaries are never set in advance,” and named the exact practice this page is about: “Some organisations have had their member schools sitting their own mock exams and have provided ‘results’ and ‘grade boundaries’ on the basis of that exercise. That’s really helpful, yes? Actually, no.” A mock grade boundary, wherever it comes from, is not an exam board figure wearing a different label. It is a number the school built, and the next three sections cover exactly how.

Which Grade Boundary Does Your Child’s School Actually Use for a Mock?

Four schools’ own published policies, read directly rather than taken from a summary, show four different answers to this question — which is itself the point. There is no single convention.

Nether Stowe School, a secondary in Lichfield, Staffordshire, sets out its method in its published Appendix A assessment policy, repeated across its Science and Design & Technology departments: “For Yr10-Yr13 only questions from past papers will be used. The software Exampro will be used to develop the papers… The overall 2023 grade boundaries will be used for Yr10-Yr13 exams with approximate 5% uplift.” That is a specific, named year’s real boundary, with a stated adjustment applied on top — not a guess, but not the current year’s boundary either, because the current year’s boundary does not exist yet.

Fearnhill School in Letchworth takes a different reference point for its GCSE Music (Edexcel) department. Its published feedback policy converts a student’s Free Composition coursework mark into “an estimated grade (using the previous year’s grade boundaries).” Not a fixed year chosen once, as at Nether Stowe, but whichever year happens to be the most recent at the time — a boundary that shifts every September as the previous summer’s results are published.

Parkstone Grammar School in Poole applies the same idea to A-level Geography (AQA). Its Year 13 curriculum booklet states that “most homework tasks are research and writing for past practice essay questions which are marked and graded to past A level mark schemes and grade boundaries. This gives a good idea of progress over the year and realistic feedback.” Here the past boundary is not reserved for one formal mock sitting; it is the department’s everyday reference point for feedback across the year.

A fourth pattern, heard directly from the person setting the mock rather than read from a policy document, comes from Ellie Upton, a science teacher at Chace Community School in Enfield, speaking on AQA’s own Inside Exams podcast. Describing how she talks her students through a marked mock paper, she says: “we talk all about how the mark scheme works and how, actually, 70% might be an A one year but it won’t be an A the next year.” She holds the actual grade back until that conversation has happened, precisely because the number on its own tells a student less than it looks like it does.

Which Paper Does the Mock Actually Use — and Why That Changes the Grade Too

Choosing a boundary is only half of what a school decides. The other half is which questions the mock actually asks, and Ellie Upton’s account on AQA’s Inside Exams podcast sets out the mechanics in more detail than any published policy does. Her department starts from “the secure material” on AQA’s own restricted platform, but rarely uses a whole paper untouched. “What we do is we quite often will change it, adapt it,” she explains — swapping an extended-answer question for a different one, or recombining questions from more than one past series, specifically because leaked papers circulate: “you do get people that will access that material and then they share it… you Google it, you can find it on the internet.”

That editing has a statistical consequence most parents never see. A real exam paper is built so that its opening questions are gentler and its later questions discriminate more sharply between students — the sequencing AQA’s own production teams design deliberately, as the same podcast episode describes elsewhere. Chop a paper up, drop a question, or splice in one from a different series, and that sequencing changes with it. A mock built from mixed-and-matched questions is not a shorter version of the same test; it is, statistically, a different test, even when every individual question came from a genuine AQA paper.

Nether Stowe School’s own policy confirms the same building method from the school side rather than the teacher’s side: “For Yr10-Yr13 only questions from past papers will be used. The software Exampro will be used to develop the papers.” Exampro is the same AQA question bank Ellie Upton describes drawing from — a live illustration that this is not one teacher’s personal habit but a documented, tool-supported practice across schools.

Who Marks It — and Does That Change the Grade?

The third variable the brief for this page names, and the one parents ask about least, is who is actually holding the red pen. GOV.UK’s current student guide is explicit about the real exam: “Exam scripts are marked completely anonymously — the marker doesn’t know who the student is, or what school or college they’re from.” Every marker is trained on a mark scheme, and awarding organisations “check markers’ work as they progress through their allocated scripts to make sure marking is consistent and high-quality.”

A mock does not go through that process. Ellie Upton, describing her own department’s routine on AQA’s podcast, says plainly: “the subject teacher will actually mark the exam paper” — the same person who taught the topic, in some cases the same person who wrote the questions. Nether Stowe School’s policy uses near-identical language for its own routine, describing formative checks “marked by the teacher[s].” Neither is anonymous, and neither goes through the moderation checks that flag a real examiner whose marking has drifted from the national standard.

That does not mean a mock is marked carelessly. Portal House School’s subject overview for Science sets out its GCSE Biology and Physics routine in detail, stating that mock papers sat “at Christmas (paper 1) and another mock exam after Easter” are followed by internal moderation: “CL and AC internally moderate Mock exam papers at Christmas and Easter — Biology and Physics GCSE.” That is a genuine quality check, carried out by two named members of staff — it is simply a different, smaller-scale check than the anonymous, board-run process a real script goes through.

Are Mock Boundaries Always Stricter? What the Named Evidence Actually Shows

A number of guides answer this question with a single confident direction: mocks are set harder, on purpose, to keep students from getting complacent. The named evidence gathered for this page does not support picking one direction. It shows schools choosing different methods, for different reasons, with outcomes that can run either way.

The clearest single case runs against the “always harsher” story. A modern-languages teacher, writing publicly about the year their school introduced the reformed GCSE specification, described exactly what happened when a mock was graded against a boundary that no longer matched the paper: “I myself experienced this the last time we had a specification change, when we used the outgoing boundaries and applied these to the then incoming specification. It looked like my class, and the cohort in general, were underachieving and failing.” The real result told a different story entirely: “When GCSE results came out, the cohort achieved a VA of over 1.5 — which is certainly not a failing cohort!” A borrowed boundary did not make that mock stricter in any useful sense. It made it wrong, and it happened to be wrong in the direction that alarmed the whole class.

The same teacher explains why a mismatch like that is close to inevitable around a specification change: “Grade boundaries are only set once all students have sat an exam, and are therefore awarded retrospectively.” No one, including the exam board, has a real boundary to reuse for a brand-new specification’s first mock — whatever number goes on the paper is invented for the occasion, in either direction.

Set against that, Ellie Upton’s AQA podcast colleague who refuses to run mocks at all gives the opposite worry, not a fixed direction but an accusation of arbitrariness: mock exams involve “making up grade boundaries left, right and centre.” And the four schools named in the previous section show three genuinely different methods sitting side by side — a fixed past year plus an uplift, a rolling previous-year boundary, and a boundary reused for ongoing homework rather than one sitting — none of which is inherently stricter or more generous than the others; each simply answers a different question about what the school is trying to measure.

The honest summary is the one this page has tried to give throughout: whose practice produced which grade determines what that grade means, and there is no substitute for asking the specific school which method it used.

Why This Year’s Real Boundary Could Never Have Been Known When the Mock Was Set

Even a school acting in good faith cannot match a number that does not exist yet. GOV.UK states that “awarding organisations set grade boundaries after exam papers or assessments have been marked” — not before, and not provisionally. AQA’s own phrasing is blunter still: “we can’t decide grade boundaries until we have seen how students have performed in the exams.” A mock sat in November or January is, by definition, measured against a boundary that will not be calculated until national scripts are marked the following summer.

That is not a technicality. It is why Ofqual has, on the record, adjusted upper-grade GCSE maths boundaries within the last month based on national performance evidence gathered over several years — a live 2026 illustration of exactly the process a school cannot replicate at its own scale, covered in full in our breakdown of how the 2026 grade boundaries were actually set. A school choosing to reuse last year’s real boundary, as Nether Stowe and Fearnhill both do, is choosing the closest available proxy for a number that is, by design, unknowable in advance — not a mistake, but a substitution the reader is entitled to know was made.

What This Means for Reading Your Child’s Mock Grade

Given all of the above, the useful response to a disappointing — or reassuring — mock grade is not to treat the letter on the page as a verdict. Three practical steps follow directly from what the four named schools and the AQA podcast actually show.

Ask which method the school used. A grade built on a fixed past year with a stated uplift, like Nether Stowe’s, means something different from one built on the rolling previous year, like Fearnhill’s, or one used all year for coursework feedback, like Parkstone’s. Many schools will simply tell a parent who asks directly, as the policies quoted on this page show they are willing to put in writing already.

Look at the raw mark and percentage, not only the grade letter. Ellie Upton’s own habit — talking through the mark scheme with her class before revealing the grade — exists precisely because the grade label depends entirely on where the school drew its line. The underlying percentage is the more stable number to track between one mock and the next.

Do not compare a grade across two different schools’ mocks. Because the boundary is chosen locally, not nationally, a grade 6 at one school and a grade 6 at another are not the same claim about ability — only a real exam boundary, set once for the whole country, supports that comparison.

None of this is about how to revise for the next mock; our separate guide to preparing for GCSE mocks covers that ground properly. It is worth knowing the actual mock dates for the current cycle too, which our guide to when GCSE mocks actually happen in 2026-27 sets out school by school. And a mock grade is a different thing again from the predicted grade a teacher eventually submits for college or UCAS purposes — that number folds in classwork, homework and professional judgement built up over the whole course, not one sitting graded against a borrowed line.