A record 146,553 candidates entered Ireland's 2026 State examinations, a 4% rise on the 140,457 who sat them in 2025. The written papers ran from 3 June to 23 June. Nearly every one of those candidates revised. Most of them revised badly, because the standard advice is not supported by the evidence.
What is exam revision?
Revision is the second pass. Learning is the first encounter with material. Revision is the deliberate retrieval and reconsolidation of material you have already met, timed so that it survives until the exam.
That distinction is not pedantic. It decides what counts as a legitimate revision activity.
If a session involves reading material for the first time, that is learning, and it belongs in term time. If it involves pulling material back out of memory under something like exam conditions, that is revision, and it belongs in the weeks before the paper. Most students who say they are revising are actually re-learning, which is why the hours feel long and the recall stays thin.
Revision methods actually work
In 2013, Dunlosky, Rawson, Marsh, Nathan and Willingham reviewed ten common study techniques for Psychological Science in the Public Interest and graded each one for utility across subjects, ages and test formats. It remains the most cited audit of its kind. Two techniques earned a high rating. Five earned a low one.
The single cleanest demonstration comes from Roediger and Karpicke's 2006 experiment. One group studied a passage four times. Another studied it once, then tested themselves on it three times. Five minutes later, the rereading group performed better. A week later, the position reversed hard.
Same material. Same total time. A 21-point gap, produced entirely by the decision to close the book. This is the mechanism behind active recall, and it is the reason a revision session that feels comfortable is usually a wasted one.
Rereading feels productive because fluency masquerades as knowledge. You recognise the page. Recognition is not retrieval.
How to build a revision timetable?
Build the timetable backwards from the exam calendar, not forwards from today. Six steps.
- Print the official timetable. Not a summary of it. The actual dated document, with morning and afternoon sessions marked. Everything below depends on it.
- Audit your gaps against the syllabus, not your notes. Take the official topic list and rate each item red, amber or green. Notes flatter you. Syllabi do not. Sia Note will compress a long chapter into a single concept card with the risk points marked, which is useful for building the red list quickly.
- Order subjects in reverse timetable order. Papers sitting in the final week get their first serious pass first, because they have the longest gap to survive. Papers in week one get the last pass, because they need to be warm on the day.
- Assign topics, not hours. A session is defined by what you will attempt to recall, not by how long you will sit there. "Tuesday 7pm, 90 minutes of Chemistry" is not a plan. "Tuesday 7pm, closed-book recall of electrochemistry, then the 2024 question on it" is.
- Schedule the second and third pass immediately. When you place a topic, place its repeats in the same action. Flashcard decks built on FSRS scheduling do this arithmetic automatically and pace the repeats against your exam date.
- Finish with full papers under timed conditions. A minimum of three per subject, marked against the official marking scheme. Mock Exam mode reproduces the format and grades the response with the reasoning attached, which is the part a marking scheme alone will not give you.
Here is step three applied to the front half of the 2026 Leaving Cert timetable. The principle transfers to any exam calendar.
Most students invert this. They revise their first exam first, then discover in week three that a subject they last touched in April has gone cold. Our guide to building a study schedule walks through the same arithmetic for semester finals.
How many hours of revision is enough?
The hour count is the wrong unit. Nobody in the research literature measures revision in hours, because two students can log identical hours and retain wildly different amounts depending on how the hours are distributed.
The useful number is the gap between repeats.
Cepeda and colleagues found in 2008 that the optimal spacing gap runs roughly 10% to 20% of the interval you need the material to survive. If your exam is 10 weeks away, the productive gap between two passes on a topic is about one to two weeks. If it is four weeks away, the gap compresses to three to six days.
That converts directly into a rule. Count the weeks to your paper, take a tenth, and never revisit a topic sooner than that unless you failed to recall it. Our breakdown of spaced repetition covers the underlying forgetting curve, first mapped by Ebbinghaus in 1885 and unbroken since.
Cramming is what happens when the gap collapses to zero. It works, briefly, and the evidence on what you can memorise the day before an exam is not encouraging beyond about 48 hours.
When do Irish exams actually fall?
Ireland compresses more exam pressure into a shorter window than any other English-speaking market. The Leaving Certificate is taken by roughly 60,000 students each year, and it is the sole gateway to CAO points. University resits then land in the summer, which catches students out.
The resit line matters more than it looks. At UCC, students who fail a module in the Winter, Spring or Summer sessions are registered automatically for the August supplemental session. Nobody has to opt in, and every year some students revise as if the deadline were September.
If you are sitting a resit, you have weeks, not months. Collapse the plan: syllabus audit, three timed past papers, and retrieval only on the topics you failed. Feed the module's lecture notes and readings into AskSia's Multi-source Q&A so answers come back with the source passage attached, rather than re-reading 200 pages you already failed to retain once.
What goes wrong in the final fortnight?
Three failure modes account for most of the damage, and all three feel like hard work.
The first is the memorised essay. Examiners see thousands of scripts and recognise a pre-learned answer that has been bent to fit the question. Learn argument structures and quotations. Do not learn essays.
The second is treating the written paper as the whole assessment. Orals, practicals and coursework components are already banked before June. A student who underperformed on a 40% oral needs a different written strategy from one who did not, and most never do that arithmetic.
The third is mock-score panic. A mock is a diagnostic, not a prediction. The correct response to a bad mock is to read the marking scheme, find where the marks actually went, and rebuild the red list. Running the topic list through AskSia's Concept Map shows which weak topics are load-bearing for later ones, which is not obvious from a syllabus printed in linear order.
Frequently Asked Questions
What is an exam revision?
Exam revision is the structured second pass over material you have already learned, timed so that recall survives to the exam date. It is distinct from learning, which is the first encounter with content and belongs in term time. The distinction has practical force: any session where you are meeting material for the first time is not revision, and treating it as such is why students report 40-hour weeks with little retention gain. The activities that qualify are retrieval-based. Closed-book recall, past papers, self-quizzing and spaced flashcard review all count. Rereading and highlighting, rated low utility by Dunlosky et al. in their 2013 review of ten techniques, largely do not. Start by separating your topic list into "not yet learned" and "learned but cold", and revise only the second group.
What is the best way to revise for exams?
Practice testing, applied across spaced sessions. These are the only two of ten techniques that Dunlosky and colleagues rated as high utility in 2013, and the effect size is not subtle. In Roediger and Karpicke's 2006 experiment, students who studied a passage once and self-tested three times recalled 61% of it a week later. Students who read it four times recalled 40%. Identical material, identical time. The practical form is: close the book, write down everything you can recall about a topic on a blank page, then check what you missed and revisit only the gaps. Repeat the topic after a gap of roughly 10% of the time remaining to the exam. Build the whole sequence around past papers rather than notes, and mark yourself against the official marking scheme.
What is a revision test?
A revision test is any self-administered assessment used as a study tool rather than a grading instrument. It covers closed-book recall, timed past papers, flashcard runs and end-of-topic quizzes. The mechanism is the testing effect: the act of retrieving information strengthens the memory trace more than re-exposure to the same information does. The format matters less than the closure of the book. A test you mark yourself, honestly, against the syllabus is worth more than an hour of re-reading, and the 21-point recall gap in the 2006 Roediger and Karpicke data is the reason. Aim for at least three full timed papers per subject in the final month. AskSia's Mock Exam mode generates practice in the real exam format and returns the reasoning behind each mark, which a bare marking scheme will not.
How many hours of revision for an exam?
Hours are the wrong unit, and no serious study in the literature reports them as the primary variable. Two students can log 100 hours each and retain very different amounts depending on how those hours are spaced. The number that predicts retention is the interval between repeats. Cepeda and colleagues found in 2008 that the optimal gap sits at roughly 10% to 20% of the interval the material must survive. Ten weeks to the exam means a one-to-two-week gap between passes on a topic. Four weeks means three to six days. Set the gap first, then let the hours fall out of the topic count. If you want an hour figure anyway, most students find that three or four focused retrieval sessions per subject per week, of 45 to 60 minutes, is sustainable across an eight-week run.
When should you start revising for exams?
Work backwards from the paper date, not forwards from today. For a three-week exam window like the Leaving Cert's 3–23 June 2026 schedule, the subjects sat in the final week need their first serious pass earliest, because they have the longest interval to survive. The subjects sat on day one get the last pass. That reversal is the single highest-leverage scheduling decision available, and most students get it backwards. For university resits, the window is much tighter. DCU's 2026 resits run 27 July to 8 August, with the timetable out by 13 July, which leaves roughly a fortnight of usable revision. Get the official timetable, mark the red topics against the syllabus, and check the course cheatsheets for a condensed version of the material you need to reconsolidate.
Is rereading notes a waste of time?
Close to it, and the evidence is consistent enough that it belongs in the low-utility category alongside highlighting and summarisation in Dunlosky's 2013 review. Rereading produces fluency, and fluency feels like knowledge. You recognise the page, mistake recognition for retrieval, and stop revising a topic you cannot actually reproduce under exam conditions. The failure only surfaces in the hall. The correction is cheap: after any pass over a set of notes, close them and write what you remember before checking. If you can reproduce 80% of the structure, the topic is warm. If you cannot, it never went in, whatever the highlighter suggests. Our guide to studying for exams covers the full retrieval workflow, and the test-prep hubs map the format of each major exam.
What can revision not fix?
Revision consolidates what you learned. It does not create what you never learned.
A topic you skipped in November is not a revision problem in May. It is a learning problem, and it needs first-pass teaching, not retrieval practice. Students routinely mistake one for the other and burn a fortnight of retrieval sessions on material that was never encoded. The honest triage is to mark those topics separately and decide, deliberately, whether to learn them from scratch or write them off and defend the rest of the paper.
The evidence base here is also narrower than it looks. Most of the retrieval-practice literature was built on verbal material and factual recall. It transfers well to definitions, dates, mechanisms and structured argument. It transfers less cleanly to open-ended design work, extended lab reports and studio assessment, where the binding constraint is time on task, not memory.
