CAT 2026 is on 29 November. At this point the syllabus is no longer the variable — your scoring behaviour is. What to change, and what to leave alone.
Short answer: treat the next three weeks as a behaviour problem, not a syllabus problem. Your coverage is what it is. What is still movable is which questions you attempt, how quickly you quit the ones you shouldn't have started, and whether your attention holds for the full duration at the hour your slot actually falls.
CAT 2026 is on 29 November 2026, conducted by IIM Indore. If you are reading this in early November, you have roughly three weeks. That is a specific and slightly awkward amount of time: too short to build a new strength, long enough that how you spend it genuinely changes your score.
Most advice at this stage is some version of "revise everything." That is not a plan, it is a description of anxiety. Here is what the three weeks are actually good for.
A topic you learn in week one of November will, by 29 November, be in the state where you recognise the question but cannot execute it quickly. Under sectional timing that is worse than not knowing it at all, because recognition tempts you into starting. You spend ninety seconds confirming you can't finish, and those ninety seconds were worth two questions you could have answered.
The exception is narrow: a small, self-contained gap in an area that shows up often, where closing it is a day's work. If it needs a week, it needs to wait for next year.
What is genuinely movable in three weeks is selection — the decision about which questions to attempt at all. Most candidates lose more marks to questions they should never have opened than to questions they got wrong. That behaviour is trainable, and three weeks is about the right length of time to train it.
Three marks for a correct MCQ answer, minus one for a wrong one, nothing for leaving it. TITA questions — where you type the answer rather than choose it — carry no negative marking at all.
Read that as an instruction rather than a rule. On MCQs, a genuine coin-flip between two options is a negative-expectation bet, so leaving it is correct. On TITA, there is no downside to attempting anything you can parse, so an unattempted TITA you could have read is a pure loss. A surprising number of candidates get this backwards under pressure: they guess on MCQs because the options feel narrowable, and skip TITAs because they feel harder.
A mock you have not analysed teaches you almost nothing. It gives you a score, which you already suspected, and a few hours of fatigue. Two mocks a week fully dissected will move you further than five skimmed.
Dissected means: for every question you got wrong, which of these was it — didn't know the concept, knew it and misread, knew it and ran out of time, or shouldn't have attempted it. Those four failure modes have four different fixes, and only the first one is about the syllabus. If most of your losses are in the other three, more revision will not help you, and three weeks of revision will feel productive while changing nothing.
Write your mocks at the time of day your actual slot falls in. Attention is not constant across the day, and you want that curve rehearsed rather than discovered on 29 November. There is more on structuring this in how to use CAT mocks properly.
The sections come in a fixed order with their own clocks. Time you don't use in one does not carry forward. If you built the habit of banking time somewhere, that habit has nothing to do here.
What you can still decide in advance is the order you work within a section, and the point at which you abandon a question. Pick both now, on paper, while calm. Decisions made at minute thirty of a section you are losing are not good decisions.
CAT runs in three slots on the same day, and the slots are not identical in difficulty. Your raw marks are normalised against the statistics of your own slot to produce a scaled score, and your percentile is computed from that scaled score — not from the raw marks.
The practical consequence is that a lower raw score in a harder slot can finish ahead of a higher raw score in an easier one, and that you cannot calibrate your own performance against someone else's raw count from a different slot. On exam evening, do not let someone else's "I got 60 marks" mean anything to you. The full mechanism is in how CAT percentile is actually calculated.
Reporting time, centre address, permitted and prohibited items, and the ID you must carry are stated on your own admit card and on iimcat.ac.in. These vary by centre and change between cycles. Read them yourself, the week before, and do not rely on a summary from any site — including this one.
Taper. Fatigue on 29 November costs more than a few extra hours of practice buys. Keep one or two short, familiar sessions a day to stay warm, sleep on the schedule your slot requires, and do a dry run of the journey to the centre if it's unfamiliar.
Resist the urge to write a mock on the last two days. A bad score will not teach you anything you can act on, and it will take something out of you that you need.
One thing worth doing in these three weeks that has nothing to do with preparation: find out what your target institutes actually weigh at final selection, and what their application timelines look like. Several CAT-accepting institutes have their own forms with their own deadlines, and some close before results are out — which means waiting for your percentile is a decision, and usually the wrong one. The applications that close after you write CAT covers that, and what IIM shortlists actually use covers the weighting.
You do not need to act on any of it this month. You do need to know it exists, so that 30 November isn't the day you find out.
Stop adding topics and start fixing how you score on the ones you already know. Three weeks is enough time to change your selection behaviour - which questions you attempt and which you leave - and not enough time to learn a new area well enough to answer it under time pressure. Most of the gain left on the table is in accuracy and question selection, not coverage.
Generally no. A topic learned this late is usually recognised but not automated, which is the worst state to be in during a timed section: you spend ninety seconds confirming you can't finish it. The exception is a small, self-contained gap in a high-frequency area, where closing it is a day's work rather than a week's.
Fewer than most people give, analysed far harder. A mock you have not analysed produces almost no learning, so two mocks a week fully dissected beats five skimmed. Keep writing them at the time of day your real slot falls in, so your attention curve is rehearsed rather than discovered on the day.
You cannot choose it - the sections appear in a fixed order with sectional timing, so you cannot carry unused time forward. What you can decide in advance is the order you attack questions within each section, and the point at which you abandon one. Deciding that beforehand is what stops a single question from eating a section.
Three marks for a correct answer and minus one for an incorrect one on multiple-choice questions, with nothing deducted for a question left unattempted. TITA questions, where you type the answer, carry no negative marking. That asymmetry is the whole argument for leaving MCQs you are unsure of and attempting every TITA you can read.
Not on its own. Mock scores are noisy and the difficulty is not calibrated to the real paper, so a single reading tells you very little. The trend across your last several mocks, and the breakdown of where the marks leaked, both matter. Panic-driven changes to a working method in the final fortnight cost more than a bad mock does.
A verified alumnus from your target institute can tell you what the last three weeks looked like for them, and what they would not repeat. Not a coach's framework — one person's actual experience, honestly described.
Find a Real One