How to Revise for an Exam in One Week
Abdulrahman YunisAI Engineer
Seven days is enough to pass and not enough to cover everything. Here is a day by day plan built around retrieval instead of rereading.
Spend day one finding out what you do not know, days two to five fixing the highest value gaps, day six on a timed paper, and day seven on light review and sleep. Do not start by reading your notes from page one.
One week is enough to pass most exams and not enough to cover everything, so the entire plan is about choosing well rather than working harder.
Day one: diagnose, do not revise
This is the day people waste, and wasting it costs more than any other single decision.
The instinct is to open the notes and start reading from the beginning. Resist it. You do not yet know which parts you are weak on, so you will spend the day on material you already knew, because that material is more pleasant to read.
Instead: take a past paper, or forty mixed questions, under timed conditions, closed book. You will do badly. That is the point. What comes out is a ranked list of what is actually broken, which is the thing you need before you can plan anything.
Then write the list of topics, mark each one red, amber or green, and note the mark value of each on the paper. Red topics worth a lot go first. Green topics worth little do not get touched this week.
Days two to five: fix, in priority order
Four days of work, and the rule for each session is the same: retrieve first, read second.
For each topic, close the notes and write down everything you can produce. Then open the notes and fill the gaps. Then close them again and redo it. Reading the chapter and hoping it sticks is the version of this that does not work, and it is the version almost everyone defaults to under time pressure.
This is not a stylistic preference. Practice testing and distributed practice were rated the two highest utility techniques in Dunlosky and colleagues’ 2013 review of ten study strategies, with rereading and highlighting near the bottom. With seven days available, spending them on the bottom of that list is expensive.
| Day | Focus | Output by end of day |
|---|---|---|
| 1 | Diagnostic paper, triage | A ranked red, amber, green list |
| 2 to 3 | Red topics, highest marks first | Each one retrieved cleanly at least twice |
| 4 | Remaining reds plus the strongest ambers | Shrinking red list |
| 5 | Mixed practice across everything covered | Questions, not notes |
| 6 | Full timed paper, then analyse | An error list, not a score |
| 7 | Light retrieval, error list, sleep | Nothing new |
Revisit everything you fix
A topic fixed on day two and never touched again will have faded by day seven. The whole week is short enough that spacing feels unnecessary and long enough that it is not.
So end each day with ten minutes on something from two days ago. It costs almost nothing and it is the difference between knowing six topics on the day and knowing six topics at the start of the week.
Day six: the timed paper
Sit a full paper under real conditions. Phone away, no notes, no pausing.
Then mark it strictly and write one line per lost mark explaining why. The reasons will fall into categories: did not know it, knew it but could not retrieve it in time, misread the question, ran out of time. Only the first is a content problem, and the other three are usually worth more marks than people expect.
Day seven is for acting on that list, which is a much better use of the last day than opening a chapter you have not read.
Day seven: resist the urge to cram
The last day should feel light, and that will feel wrong.
Review the error list. Do a short retrieval session on your weakest red topics. Pack your things. Sleep properly.
New content learned on day seven is unreliable under pressure and it displaces sleep, which is when the week’s work consolidates. The marginal topic is worth less than the alertness you trade for it.
What to do when a week is genuinely not enough
Sometimes the honest arithmetic says you cannot cover the syllabus. Triage harder rather than faster.
Find what appears on every past paper. Find the core concepts that other topics depend on, since understanding those often lets you attempt questions on material you never revised. Cover those properly and accept gaps elsewhere.
A broad competent pass across the whole paper usually scores better than depth in one corner, because mark schemes reward basics generously and depth sparingly.
If you have several exams in that week rather than one, the plan changes shape, and that is covered in how to study for multiple exams at once.
Where a tool fits
The slow part of this plan is day one. Building a diagnostic set from your own material, by hand, can eat the morning you needed for the diagnosis itself.
That is what StudyLabAI’s exam practice generator is for. Upload the lectures or notes and get a timed set from that specific material, so you can run the day one diagnostic in an hour instead of building it. It handles the construction. The triage decision, and the discipline of retrieving before reading, is what actually earns the marks.
Common questions
Is a week enough time to revise properly?
Enough to pass most exams and to do well on ones you have attended through the term. Not enough to learn a module from nothing, and pretending otherwise leads to covering everything badly.
How many hours a day should I do?
Six to eight with real breaks, and no more than four of those on genuinely demanding retrieval work. Twelve hour days in a revision week produce exhaustion by day four.
Should I make notes during the week?
Only very short ones, and only as an output of retrieval. Making a fresh set of notes from the textbook is the single most common way a revision week disappears.
What if I get a bad score on the day one paper?
Expected, and useful. The score is not a prediction, it is a map. A bad day one score with six days left is a much better position than a comfortable feeling and no information.
Should I revise with other people?
For testing each other, yes, it works well. For sitting in a library together reading separately, it mostly adds social pressure without adding retrieval.