Updated 21-September-2026
If the exam is tomorrow, most advice arrives too late. "Start revising early" and "space it out over weeks" are correct, but useless to a student who is starting now. This piece is about what to actually do when the runway has already shrunk to a day or two, based on what the underlying memory and sleep research says happens in that specific window, not generic study tips repeated from a syllabus checklist.
The honest answer is more specific than "cramming is bad." A widely cited meta-analysis by Cepeda, Pashler, Vul, Wixted and Rohrer, reviewing decades of distributed-practice research, tested exactly this question: does massed practice (cramming everything into one sitting) ever catch up with spaced practice (splitting the same material across separate sessions) when the test is coming up soon? Their answer, even at the shortest retention intervals tested, was no. Spaced presentation outperformed massed presentation across every interval examined, including intervals of under a minute, and the researchers found no evidence for a "massing benefit" at short delays, only at intervals of a few seconds, which has no practical relevance to exam revision.
The practical takeaway is not "don't bother, it's too late." It's that even inside a genuinely tight window, structure still matters. Two three-hour blocks with a gap between them, on the same topic, will produce stronger recall on exam day than one six-hour block, because the gap forces the brain to reconstruct the material rather than simply hold it in working memory. If a student has 48 hours left and four subjects to cover, the better structure is short, separated sessions per subject repeated across both days, not one exhaustive pass through each subject back to back. This is a genuinely different instruction from "break it into chunks and take breaks," which is close but not the same claim: the gap between sessions on the same topic is what does the work, not just the presence of breaks.
Generic revision advice says "get 7 to 8 hours of sleep" without explaining why, which makes it easy to dismiss under pressure. The actual research is more interesting and more specific. A study by Okano and colleagues tracked 88 university chemistry students using Fitbits throughout a semester and compared their sleep patterns to exam performance. The finding, summarised by McGill University's Office for Science and Society, was that sleep duration and quality on the single night before an exam showed no meaningful correlation with test scores. What did predict performance, accounting for close to a quarter of the variance between students, was sleep quality and consistency across the entire month leading up to the exam.
Two things follow from this, and both cut against the usual framing. First, one restless night before a big exam is not the catastrophe it is often made out to be. A student who sleeps badly the night before, for ordinary pre-exam nerves, is not meaningfully disadvantaged by that fact alone. Second, and this is the caveat this study's own authors would want kept front and centre, the finding is correlational, not causal. Students with better sleep habits across a month may simply be more organised and less anxious generally, which independently predicts better exam performance. It does not follow that deliberately pulling an all-nighter to cram is a good trade, because the cramming itself is already the weaker study method per the section above, and acute sleep loss still measurably impairs the kind of working memory and attention that exam performance under timed conditions requires. The conclusion is narrower than either "sleep is everything" or "one night doesn't matter": don't panic about a single anxious night, but that finding is not a licence to sacrifice sleep for extra cramming, because the extra cramming was not going to pay off anyway.
When time is short, not all revision activities are worth equal time. Ranked by the strength of the evidence behind them, for a student with realistically one to two days left:
Highest priority: timed past-paper practice under exam conditions. This combines retrieval practice with exposure to the actual question formats and mark schemes a student will face. Merit Tutors has covered the specific mechanics of doing this well in How to Use Past Papers to Revise Effectively, including how to mark against official schemes and what to do with the errors it surfaces.
Second priority: active self-testing, not passive review. Flashcards used correctly (attempting recall before checking the answer) and mind maps drawn from memory rather than copied from notes both force retrieval, which is the mechanism that builds durable recall. The distinction between using these tools actively versus passively is covered in Effective Revision Techniques: From Flashcards to Mind Maps.
Third priority: explaining topics aloud, unscripted. Talking through a concept without notes and noticing exactly where the explanation breaks down is a fast way to find gaps in a compressed timeframe, because it exposes surface familiarity that would otherwise pass unnoticed until the exam itself.
Lowest priority, and often the biggest time sink: re-reading notes or textbooks. This is the activity that feels the most productive and delivers the least. It creates a sense of familiarity with the material without testing whether it can actually be retrieved under pressure, which is the only thing the exam will measure. In a genuinely short window, time spent re-reading is time not spent on the three activities above.
Starting a new topic from scratch inside the final 24 hours is rarely worth it; the marginal benefit of touching material for the first time this close to an exam is low compared to strengthening recall on topics already covered. Equally, trying to cover every topic on the syllabus with equal weight ignores that exam boards do not weight every topic equally either; time spent confirming which topics carry the most marks, through the syllabus or a tutor, is rarely wasted even at this late stage. For students whose anxiety is the bigger obstacle to a productive final day rather than a lack of technique, How to Manage Homework Stress: A Parent's Guide to Supporting Your Child and Top Tips for Stress-Free SATs Preparation cover the underlying anxiety management in more depth than is useful to repeat here.
Is it too late to revise the night before an exam?
No. One evening of focused, active revision, structured around self-testing and past papers rather than re-reading, still measurably improves recall over doing nothing. It will not replace weeks of preparation, but it is not wasted effort.
Should I pull an all-nighter to fit in more revision?
No. The evidence on cramming shows it underperforms spaced study even at short notice, so the extra hours from an all-nighter buy less than they appear to, while acute sleep loss still impairs the attention and working memory an exam requires.
What is the single most useful thing to do with two days left?
Timed past-paper practice under real exam conditions, marked against the official scheme, is the highest-value use of limited time because it combines active recall with direct exposure to how the exam will actually test the material.
Does one bad night's sleep before an exam ruin performance?
Not on its own. Research tracking students' sleep across a full semester found the month of sleep leading up to an exam mattered far more than the single night before it. One restless night from nerves is not the setback it is often assumed to be.
Sources:
Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2006). Distributed Practice in Verbal Recall Tasks: A Review and Quantitative Synthesis. Retrieved via ERIC: https://files.eric.ed.gov/fulltext/ED536925.pdf
Recent Posts
Categories
Tags
