EPQ Examples: What A* Projects Look Like on Each Board

By Michael Thompson · Education Specialist; 10 years teaching the IB at Bromsgrove School · Published 24 October 2026 · Updated 25 October 2026

EPQ examples are harder to find than students expect - not because good projects are rare, but because the boards publish their exemplar material mainly for teachers. AQA's exemplar materials and commentaries sit on Centre Services, its secure teacher site; OCR's specification points teachers to 'candidate exemplars with commentaries' alongside its teaching guides; Pearson keeps an exemplar library on its public website. What this guide does instead is show you the shape of an A* project, criterion by criterion, using worked outlines written for this guide, short quoted phrases from the official specifications, and the specific warnings that OCR moderators published after the summer 2024 series. Read this alongside your board's specification and you will know exactly what evidence a moderator is looking for - and what separates a top-band project from a mid-range one.

Key Takeaways

• Official exemplars are aimed at teachers: AQA's exemplar materials and commentaries are on its Centre Services teacher site, OCR's specification points teachers to candidate exemplars with commentaries, and Pearson has an exemplar library on its website - so ask your supervisor what your board provides, and treat any other 'EPQ example' you find online as a school's own resource or an illustration, not an official standard.
• AO3 carries the most marks on every board: Developing and realising the project is worth 40% on AQA (20/50 marks), 44% on Pearson (24/54 marks), and 40% on OCR (24/60 marks) - making the quality of the project product itself the single biggest grade driver.
• OCR moderators identified a clear difference between good and weak projects: Candidates who did well selected a focused topic, narrowed their research as they progressed, and evaluated with genuine self-reflection; those who did less well stuck to generic topics, cited URLs instead of formatted references, and produced descriptive evaluations.
• Every project type requires a written report: AQA requires approximately 5,000 words for a written product and a minimum 1,000 words when the product is an artefact; OCR sets the same 1,000-word minimum for the written piece that accompanies a non-written outcome; Pearson's specification gives no word count, so take the figure from your unit guidance and tutor-assessor.
• A focused research question is the foundation: OCR's 2024 moderators' report flags that projects were sometimes 'too limited or too large in scope' for 120 guided learning hours, and that candidates who did well refined a research question that was targeted and focused.
• Using examples risks stifling your own project: OCR moderators noted that templates 'can also stifle creativity and independence' when candidates do not adapt them - treat any example as a structural reference, not a script.

In This Article

1. What an EPQ example actually is (and where boards keep them)
2. How the marks are split: AQA, OCR and Pearson compared
3. What the A* grade looks like: AQA's own descriptors
4. What OCR moderators said separates strong EPQ projects from weak ones
5. Worked outline 1: an A*-shaped EPQ dissertation
6. Worked outline 2: an artefact EPQ project and outline 3: an investigation
7. EPQ title and research question examples across subjects
8. How to use EPQ examples without compromising your own project
9. What to do next

1. What an EPQ example actually is (and where boards keep them)

If you search for EPQ examples online, you will find student essays on revision websites, PDFs on school intranets, and forum threads where people have uploaded their own work. None of that is what the exam boards mean by an exemplar. AQA's exemplar materials and commentaries are on Centre Services, its secure teacher site, offered 'to help you feel confident in your assessment of project work', so students usually see them only when a teacher shares them. OCR's specification tells teachers that 'a number of candidate exemplars with commentaries' are available alongside its teaching guides, and Pearson runs an exemplar library on its website organised by unit and topic. Whatever your board provides, the material exists to support consistent marking, not to hand students a template - and the OCR moderators' report warns that templates 'can also stifle creativity and independence'.

What this guide provides instead is the shape of an A* project, described criterion by criterion, using the boards' own published specifications. Section 2 compares how AQA, OCR and Pearson split marks across their four assessment objectives. Section 3 quotes short phrases from the grade descriptors exactly as AQA publishes them in its Level 3 EPQ specification (7993). Sections 5 and 6 contain worked outlines written for this guide and labelled as such: they are not student submissions and they are not board exemplars.

One non-obvious risk is worth naming early. An EPQ essay example found on an open website may itself contain errors, a poorly structured argument, or a bibliography that does not meet the formatting standard OCR moderators expect. Copying the structure of a flawed project compounds the original problem. There is also an authenticity issue: the EPQ must be the student's own work, and examiners who recognise a circulating template will question whether the thinking behind it is original.

The guide uses that access gap productively. Where boards publish criteria, descriptors, and reporting language in documents you can read yourself, this guide quotes and links them directly, so you can see what a strong project actually satisfies, not just what one looks like on the surface.

2. How the marks are split: AQA, OCR and Pearson compared

All three boards use the same four Assessment Objectives, but the mark totals and weightings differ. The table below shows the verified figures.

Sources: AQA 7993 specification v1.4 / OCR H857 specification v2.1 / Pearson ZPJ30 specification Issue 4

AO3 is by far the biggest single block on every board, at 40-44% of the total. That means the quality of your actual project product and the research journey that produced it matters more than your log entries alone. A well-kept log that accompanies a thin project will not rescue your mark.

A few structural differences are worth noting before you pick a board or a route:

• Pearson offers four named units: P301 Dissertation, P302 Investigation/Field Study, P303 Performance, and P304 Artefact. Each unit maps onto the same AO framework.
• OCR permits: a report, a dissertation (approximately 5,000 words), an artefact, or a performance or event. Any non-written outcome must be accompanied by a written piece of at least 1,000 words, and students submit a Project Progression Record (PPR) covering the full project arc.
• AQA requires a written report for every project type. A solely written product should be approximately 5,000 words; an artefact product requires a minimum 1,000-word written report alongside it.

On UCAS points, the OCR specification says the EPQ is 'worth up to 28 UCAS points in the new tariff', and Pearson's guidance for students says 'the grade you receive is worth up to 28 UCAS points'. The same Pearson guidance says students have 120 hours to complete their project.

The practical takeaway: whichever board your school uses, plan your time around AO3. The log and review sections matter, but the project itself is where the largest single share of the marks sits.

3. What the A* grade looks like: AQA's own descriptors

AQA's marking descriptors are the closest thing to an official EPQ exemplar that most students ever see. Reading them against the lower-grade descriptors is more useful than any sample essay, because they tell you exactly where marks are won and lost.

The AQA specification describes A* performance across all four objectives. The phrases that appear at the top band include: 'effective design and thorough plan', work completed in a 'highly organised fashion', use of 'a wide range of resources critically', ability to 'fully explore the complexities of the topic', and 'full, in-depth evaluations'.

At grade C, the plan 'enables them to achieve the overall objectives', students 'use a range of resources, analyse data and apply findings' and 'explore some complexities of the topic', and they 'reflect effectively' in their evaluations. At grade E, the plan is merely 'workable', students 'use a limited range of resources' and 'analyse some of the data', and they 'give some thought' to how the outcome emerged. The gradient is consistent: every step down reflects less depth, not a different kind of work.

What do the top-band phrases mean in practice?

• 'Highly organised fashion' points directly to your Production Log. The AQA June 2023 examiner report flags that a minority of logs read as retrospective diaries rather than contemporaneous records. A log written after the fact cannot show real-time decision-making, which is precisely what AO1 rewards.
• 'Wide range of resources critically' means evaluating each source's methodology, bias, and limitations, not assembling a long bibliography. Listing fifteen sources without comment sits in the lower bands.

The non-obvious gotcha: the examiner report notes that quantitative-only monitoring of progress, tracking word count alone for instance, is not enough for top-band AO1. The log must show qualitative reflection on decisions, not just milestones reached.

4. What OCR moderators said separates strong EPQ projects from weak ones

The OCR H857 moderators' report for summer 2024 is one of the most direct accounts of what actually separates a strong project from a weak one. The patterns are consistent enough to be useful as a checklist before you start, not just after you finish.

What candidates who scored well did

• Chose a topic they genuinely cared about. Moderators noted that successful candidates 'selected a topic which genuinely interested them' - enthusiasm tends to produce the sustained effort that 120 guided learning hours demand.
• Built a focused question or brief. Strong projects had a 'targeted and focused' research question (or, for artefact projects, an equally specific brief) rather than a broad theme dressed up as a question.
• Planned carefully and adapted deliberately. High-scoring candidates 'planned and managed work carefully and painstakingly' and, crucially, 'made and justified changes to their plans and topics' when circumstances shifted. Changing course is fine; changing course silently is not.
• Referenced properly. Well-formatted in-text citations and bibliographies appeared consistently. Hyperlinks pasted into a bibliography are not formatted references.
• Narrowed as they went. Rather than maintaining a wide sweep throughout, strong candidates 'narrowed and deepened research as they progressed.'
• Evaluated with real self-awareness. Moderators looked for 'genuine self-reflection and insight', not a summary of what was done.

What candidates who scored less well did

• Picked a generic topic and never sharpened it into a workable question.
• Kept going with plans that were clearly off track rather than revising them.
• Left out in-text citations entirely, or substituted URLs for properly formatted references.
• Researched either too broadly and shallowly, or too narrowly and shallowly, without identifying and closing the gaps.
• Wrote evaluations that described the project rather than judging it.

Two scope problems worth knowing about

The report flagged that projects were sometimes 'too limited or too large in scope' for the 120 GLH available - a reminder that scope is a design decision, not something to leave until the PPR review. The moderators also specified that 'at least some aims should relate to the steps needed to achieve planned outcomes', so a list of aims focused entirely on skills development will not satisfy the criterion.

On portfolios: moderators 'want to see a curated portfolio of evidence', not every document produced. For research projects, the practical implication is that surveys should only be included 'where they can meaningfully add to the outcome' - running one to pad the portfolio is unlikely to help.

5. Worked outline 1: an A*-shaped EPQ dissertation

This outline is written for this guide. It is not a real student submission and not an exam-board exemplar. It shows the shape an A*-level dissertation project takes, mapped against AQA's four assessment objectives.

The title and research question

Title: Should the UK introduce a statutory duty of care for social media companies towards users aged under 18?

Research question: To what extent does existing UK tort law provide adequate protection for adolescent users of social media platforms, and what would a statutory duty of care add?

This title works because it is anchored to a specific jurisdiction, a specific age group, and a specific legal mechanism. A weaker version, "Is social media harmful to teenagers?", would leave the argument without a clear analytical boundary.

Five-chapter structure

Source mix

• Peer-reviewed journals: legal and public health scholarship on adolescent online harm
• Official statistics and legislation: Ofcom data, Online Safety Act 2023, Law Commission reports
• Case study: a named ruling where duty-of-care arguments were tested in court
• Counterargument material: platforms' published community standards and industry-body submissions to parliament

The annotated bibliography, which sits under AO2, notes each source's reliability and flags gaps, for instance where no longitudinal data yet exists for the post-Online Safety Act period.

Mapping to the assessment objectives

AO1 (Production Log, 10 marks under AQA): The AQA June 2023 examiner report makes a point worth noting: high AO1 marks require evidence of detailed project planning, not solely essay planning, and quantitative-only monitoring (word counts, hours logged) was flagged as insufficient for the top band. The Production Log should record the formal proposal, changes to the research question and the reasons for them, advice received and acted on, and a review of the completed project. Per the AQA specification, all of that belongs in the Production Log, not in the dissertation itself.

AO2 (Use resources, 10 marks): The annotated bibliography demonstrates critical engagement: why source A was chosen over source B, where a source is authoritative but potentially dated, where a gap in the literature became a finding in itself.

AO3 (Develop and realise, 20 marks): The dissertation runs to approximately 5,000 words, as the AQA specification sets for a research-based written report. The analytical thread runs from the research question through every chapter: each section advances the argument rather than summarising a new topic.

AO4 (Review, 10 marks): The final evaluation judges the method. What worked (the case-study chapter gave the abstract legal question concrete grounding); what was limited (primary sources were unavailable because parliamentary submissions are anonymised); what would change on a second attempt (a broader comparative angle, looking at the EU's Digital Services Act alongside UK law).

The non-obvious gotcha here is the AO1/AO3 split. Students sometimes treat the Production Log as a formality and pour all their effort into the essay. The examiner report flags precisely that mistake: the Log is assessed independently, and reflective, qualitative entries showing why plans changed matter more to AO1 than a tidy spreadsheet of hours worked.

6. Worked outline 2: an artefact EPQ project and outline 3: an investigation

Both outlines below were written for this guide. They are not student submissions or board exemplars.

Outline 2: artefact project (example subject: product design or textiles)

The brief: Design and build a flat-pack storage unit for a small student bedroom, optimising for cost under £40 and assembly time under 20 minutes.

A focused brief like this matters because it gives the review section something to measure against. Vague briefs produce vague evaluations.

What the development record must contain

• Stage 1: annotated sketches and material costings (evidence of early design decisions)
• Stage 2: first prototype with photographs, plus a log entry explaining what changed and why
• Stage 3: revised prototype photographs with notes on structural failures addressed
• Final product: photographs of the completed artefact

Per the AQA specification, photographs of each stage of development and the final product are acceptable evidence, and large artefacts do not need to be physically submitted.

The written report: a floor, not a ceiling

The AQA June 2023 examiner report records that students incorrectly treated the 1,000-word minimum as a maximum. The report should cover: literature on ergonomics and flat-pack design, material selection rationale, and a critical evaluation of the sources that informed the design decisions. One thousand words is the starting point.

Mapping to AOs

One further point: the AQA specification describes only two shapes of project product - a research-based written report on its own, or an artefact with a written report of at least 1,000 words - so there is no separate performance route under AQA, and a performance, film or composition is treated as an artefact product.

Outline 3: investigation (example subject: psychology or biology)

Hypothesis: Increasing background noise from 40 dB to 70 dB (independent variable) reduces the number of words recalled after a 2-minute delay in participants aged 16 to 18 (dependent variable).

A stated, testable hypothesis with precise variables is what separates an investigation from a general literature review.

What the method section must justify: choice of within-subjects or between-subjects design, standardised instructions, control of confounding variables (time of day, prior caffeine intake), and sample size. The method section is where AO2 lives: the quality of sources used to justify the design is as important as the design itself.

Data collection and analysis: raw data tables, descriptive statistics, and a chart showing the distribution of scores across conditions.

Limits section: acknowledge sample size constraints, absence of random sampling, and any uncontrolled confounds. This is where AO4 starts to take shape before the formal review.

A note on surveys: OCR moderators caution that surveys should be conducted only where they can meaningfully add to the outcome. If a short questionnaire is unlikely to produce data beyond what existing literature already shows, a practical design may serve the investigation better.

Mapping to AOs

7. EPQ title and research question examples across subjects

The difference between a strong EPQ title and a weak one is scope. OCR moderators noted that candidates who performed well 'refined a research question which was targeted and focused', while those who struggled 'selected a generic topic'. A good title gives you a defined argument to make, not an infinite subject to summarise.

Question stems that work

Certain openings push you towards analysis rather than description:

• 'To what extent...' forces you to weigh evidence on both sides
• 'How far does...' signals a judgement is coming
• 'Should...' sets up an evaluative position
• 'Is X the most significant factor in...' demands comparison

'What is...' questions tend to fail. They invite a report, not an argument, and moderators can see the difference in the writing.

Twelve focused titles, written for this guide

Each title below was written for this guide as an illustration. The bracketed note explains why the scope fits 120 GLH.

Two titles that fail, and why

These are written for this guide to show what not to do.

• 'What is climate change?' - No analytical angle, no scope limit, no argument. An examiner reading this title knows the candidate will describe rather than judge. OCR moderators flagged projects as 'too large in scope' for 120 GLH; a question this open makes that almost inevitable.
• 'Is social media bad?' - The word 'bad' is undefined, the platform is unnamed, the population is unnamed, and the word 'is' signals a binary answer rather than a reasoned position. A focused version would name a platform, a metric, and a group.

The practical fix is to narrow twice. Take your broad interest, add a specific context, then add a measurable or debatable outcome. That usually gets you to a title a supervisor can approve.

8. How to use EPQ examples without compromising your own project

EPQ examples are most useful as a checklist, not a template. The distinction matters because your project must be entirely your own work, and every source you consult, including any example project, must be cited. Using another student's structure or argument without attribution risks academic misconduct, regardless of how you came across the work.

The risk runs deeper than plagiarism, though. The OCR moderators' report for summer 2024 says that while templates work in some centres where each candidate adapts them to their own project and learning style, "they can also stifle creativity and independence." The same logic applies to EPQ examples. If you follow one closely, your Project Progression Record or AQA Production Log may end up reflecting someone else's decision-making process rather than your own. That is a direct problem for AO1, which assesses how you managed the project independently.

The non-obvious gotcha: a well-written example can make your own early planning feel adequate when it isn't. Seeing a polished final product tells you nothing about whether your current proposal has gaps.

Use epq examples this way instead:

• Check that your plan covers every required element: proposal, research, development, evaluation, and presentation.
• Confirm your scope fits 120 guided learning hours, as required by AQA and OCR alike.
• Note what a strong evaluation looks like structurally, then put the example aside before drafting your own.

Once you have ticked those boxes, the example has done its job.

9. What to do next

Before you study another EPQ example, confirm two things: which board you are registered with, and which project type you are doing. The mark scheme weightings differ across AQA, OCR and Pearson, and the log has a different name on each. AQA calls it the Production Log and Assessment Record; Pearson uses an activity log; OCR uses the Project Progression Record (PPR). Mixing up the vocabulary in your submission is a small but avoidable problem.

The most useful thing you can do this week is ask your supervisor or EPQ coordinator what exemplar material your board provides and whether they can share it with you. AQA's exemplar materials and commentaries are on Centre Services, its teacher site; OCR's specification points teachers to candidate exemplars with commentaries; Pearson has an exemplar library on its website. These are marked projects, typically with the board's commentary, not illustrative outlines, and seeing how a real submission was judged tells you more than any summary can.

Ask your coordinator for board exemplars this week. Confirm your project type at the same time, then map your current plan against the correct AO weightings before your next supervision meeting.

FAQ

Where can I find real EPQ examples to read?

AQA's exemplar materials and commentaries are on Centre Services, its teacher site; OCR's specification points teachers to candidate exemplars with commentaries; Pearson has an exemplar library on its website organised by unit and topic. Ask your supervisor or EPQ coordinator what your board provides and whether they can share it with you.

How long does an EPQ dissertation have to be?

A project product consisting solely of a research-based written report should be approximately 5,000 words on AQA; OCR also specifies approximately 5,000 words for a dissertation outcome; where the product is an artefact, a minimum 1,000-word written report is required on both AQA and OCR.

What do EPQ assessors look for in an A* project?

AQA's A* grade descriptors include phrases such as 'effective design and thorough plan', 'wide range of resources critically', and 'full, in-depth evaluations'; OCR's 2024 moderators' report adds that top candidates selected a focused topic, narrowed their research as they progressed, and evaluated with genuine self-reflection.

Is the EPQ worth the same UCAS points on every board?

Yes - every board gives the same figure: the OCR specification says 'up to 28 UCAS points in the new tariff', Pearson's guidance for students says 'the grade you receive is worth up to 28 UCAS points', and AQA's EPQ page describes the qualification as 'worth half an A-level (28 UCAS points)'.

Do I need a written report even if my EPQ is an artefact?

Yes - AQA states that all project products must include a written report, and where the product is an artefact the report must be a minimum of 1,000 words; OCR sets the same 1,000-word minimum for the written piece that accompanies a non-written outcome, while Pearson's specification gives no figure, so take it from your unit guidance and tutor-assessor.

Can I use a survey as part of my EPQ?

OCR's 2024 moderators' report advises that surveys should be conducted only where they can meaningfully add to the outcome - using a survey simply to generate primary data, without it genuinely informing the project, is flagged as a weakness in that report.

References

• AQA Level 3 Extended Project Qualification (7993) Specification v1.4 - 2.5.1 Assessment evidence and project product - https://filestore.aqa.org.uk/subjects/AQA-W-7993-SP-19.PDF
• Cambridge OCR Level 3 Extended Project Qualification H857 Specification v2.2 (October 2026) - Assessment, PPR, marking criteria bands, overlap with other qualifications - https://www.ocr.org.uk/Images/414382-specification.pdf
• Pearson Edexcel Level 3 Extended Project (ZPJ30) Specification Issue 4 - Assessment objectives, the presentation, assessment evidence - https://qualifications.pearson.com/content/dam/pdf/Project-Qualification/Level-3/2010/Specification/project-specification-level-3.pdf
• AQA Level 3 Extended Project Qualification 7993 - Report on the examination, June 2023 - https://www.aqa.org.uk/files/sample-papers-and-mark-schemes.2023.june.AQA-7993-WRE-JUN23_PDF/fa246dd5d7dc686c17d93e113bdfafcea9e00317.pdf
• Cambridge OCR Extended Project H857 Moderators' report - Summer 2024 - https://www.ocr.org.uk/Images/727849-moderators-report.pdf