CAT Quantitative Aptitude Strategy: Sequence, Not Coverage
CAT QA is 22 questions in 40 minutes, and the mathematics tops out around class 10. There is no calculus, no matrices, no advanced trigonometry. Almost every candidate already knows the topic list.
What separates scores is not coverage but two other things: the order you study topics in, and whether the arithmetic underneath them is automatic. This page is about both, plus the specific reason strong engineering students often underperform here.
Recognition, not calculation
Under two minutes a question is not enough time to work through a problem methodically. It is enough to recognise the type, apply a known approach, and move. That distinction — recognition versus calculation — is what CAT quant actually rewards.
Quick Answer (30-Second Read)
Study order
Arithmetic → Algebra → Geometry
Modern maths last
Attempt range
14–18 of 22
At 80%+ accuracy
- Arithmetic first, always. It is the largest block, it compounds into everything else, and it feeds DILR caselets too.
- Memorise the arithmetic furniture — tables to 20, squares to 30, cubes to 15, percentage-fraction conversions. Unglamorous and the highest-return week you will spend.
- Leaving four questions untouched is correct. Some are deliberately built as time sinks.
- Engineers often underperform by solving rigorously rather than recognising quickly. Rigour is the wrong instinct here.
- Modern maths last. Smallest block, hardest to make fast — poor return early on.
Section structure per the official CAT information bulletin at iimcat.ac.in. Weightage below is observed from past papers, not official.
CAT Mock Test — VARC, DILR & QA on the real interface
Knowing a topic and solving it in 109 seconds are different skills. The time-per-question view shows which questions actually consumed your section.
The study sequence that compounds
Every candidate has the same topic list. Far fewer sequence it so that earlier work makes later work easier — and the difference shows up as speed, not knowledge.
| Order | Area | Topics | Why here |
|---|---|---|---|
| 1 | Arithmetic | Percentages, ratio and proportion, profit and loss, simple and compound interest, time-speed-distance, time and work, mixtures, averages | Largest block, and percentages underpin most of the rest. Also feeds DILR caselets |
| 2 | Algebra | Linear and quadratic equations, inequalities, functions, logarithms, progressions, maxima and minima | Uses ratio and percentage fluency built in step 1 |
| 3 | Geometry & mensuration | Triangles, circles, quadrilaterals, polygons, coordinate basics, solids | Largely self-contained. Formula-dense rather than method-dense |
| 4 | Number system | Factors, remainders, divisibility, LCM and HCF, base systems, last-digit problems | Self-contained, and rewards pattern recognition over technique |
| 5 | Modern maths | Permutation and combination, probability, set theory | Smallest block and slowest to become fast. Poor return if done early |
The most common sequencing mistake is starting with permutation and combination because it feels like "real" CAT maths, then reaching arithmetic in month four. Arithmetic then never gets the repetition it needs to become automatic — and arithmetic is precisely the area where automaticity, rather than understanding, decides your speed.
Percentages are load-bearing
Profit and loss is percentages. Interest is percentages. Mixtures is a weighted average of percentages. Data interpretation is percentages under time pressure. A candidate who can convert and compare percentages instantly has effectively pre-solved a large part of several other topics, which is why it sits first on the list.
The arithmetic furniture
Boring, mechanical, and the highest-return week of preparation available. Every second not spent calculating is a second available for the next question.
| What to memorise | Range | Where it pays |
|---|---|---|
| Multiplication tables | Up to 20 | Everything. This is the base layer |
| Squares | Up to 30 | Geometry, algebra, number system |
| Cubes | Up to 15 | Number system, algebraic identities |
| Percentage ↔ fraction | 1/2 to 1/16 | Arithmetic, DI, anything with ratios |
| Powers of 2 and 3 | 2 to 2¹², 3 to 3⁶ | Number system, base conversion |
| Pythagorean triples | Common ones | Geometry — saves a full calculation each time |
| Prime numbers | Up to 100 | Factorisation, divisibility, remainders |
What this buys you, concretely. A question asks for 37.5% of 4,896.
Without the conversion: multiply 4,896 by 37.5, divide by 100. Roughly 25 seconds, with a real chance of a slip.
With it: 37.5% = 3/8. So 4,896 ÷ 8 = 612, then × 3 = 1,836. About 6 seconds.
Nineteen seconds saved on one question. Across a section, that is another two questions attempted.
Candidates skip this because it feels like school work rather than CAT preparation. It is school work — and it is the difference between finishing a question in six seconds and twenty-five, repeated across an entire section.
Is your QA problem knowledge or speed?
These need entirely different fixes, and the time-per-question view is the only thing that tells them apart honestly.
- Time per question
- Topic-wise accuracy
- Sectional percentile
Why strong engineering students underperform in CAT quant
This surprises people, because engineers arrive with far more mathematics than CAT requires. The problem is not knowledge — it is that engineering trains habits CAT punishes.
| Engineering habit | Why it costs marks in CAT | What to do instead |
|---|---|---|
| Solve rigorously and completely | 109 seconds does not allow a full derivation | Recognise the type, apply the shortcut, verify roughly |
| Reach for the general method | CAT questions often have a specific trick that skips the general case | Look for the shortcut first, general method as fallback |
| Distrust approximation | Options are often far apart; exact values are unnecessary | Check option spacing before deciding how precise to be |
| Refuse to abandon a solvable problem | Solvable is not the test — solvable in 109 seconds is | Fixed per-question ceiling, enforced |
| Study advanced topics for comfort | Calculus and matrices are not tested at all | Redirect that time to arithmetic speed |
The third row is worth reading twice. CAT frequently sets options far enough apart that a rough estimate identifies the answer without completing the calculation. Engineers tend to find this uncomfortable and solve fully anyway, spending sixty seconds to confirm what fifteen seconds of estimation had already established.
None of this means an engineering background is a disadvantage. It means the habits need adjusting, and candidates who make that adjustment early tend to move fastest in QA of anyone.
In-exam triage
Forty minutes, twenty-two questions, and no obligation to attempt them in order. Triage is a skill you should rehearse rather than improvise.
| Pass | Time | What you do |
|---|---|---|
| First pass | ~20 min | Solve everything you recognise immediately. Skip anything that needs thinking about how to start |
| Second pass | ~13 min | Return to skipped questions where you now see an approach. TITA questions get priority — no penalty |
| Final pass | ~5 min | Educated guesses on MCQs where two options are eliminated. Fill every remaining TITA |
| Last 2 min | 2 min | Stop attempting. Check nothing sits in an unintended state |
The rule that makes the first pass work: if you cannot see how to start within fifteen seconds, skip it. Not "if you cannot solve it" — if you cannot see the opening move. Questions where the approach is not immediately visible are exactly the ones that consume four minutes and return nothing.
Leaving four questions untouched at the end of a QA section is a normal, correct outcome. Some CAT quant questions are built as time sinks, and recognising one is worth more than solving it.
Five mistakes that cost QA marks
| Mistake | What it costs | The fix |
|---|---|---|
| Studying topics in the wrong order | Arithmetic never becomes automatic | Arithmetic first, modern maths last |
| Skipping the arithmetic furniture | 15–20 seconds lost on every calculation-heavy question | One week on tables, squares, cubes, fraction conversions |
| Solving in the printed order | A hard question early eats the time easier ones needed | Three-pass triage, 15-second start rule |
| Calculating exactly when options are far apart | 45 extra seconds for no additional marks | Check option spacing before deciding precision |
| Leaving TITA questions blank | Free marks declined — there is no penalty | Fill every TITA, always, even with an estimate |
Key Takeaways
- Sequence beats coverage. Arithmetic first because it compounds; modern maths last because it is small and slow to master.
- Memorise the arithmetic furniture. 37.5% of 4,896 takes 6 seconds with fraction conversions and 25 without.
- 109 seconds rewards recognition, not calculation. Identify the type and apply a known approach.
- Engineers should unlearn rigour. Estimate when options are far apart, and abandon solvable questions that are not solvable in time.
- If you cannot see the opening move in 15 seconds, skip it. Those are the four-minute sinkholes.
- Leaving four questions untouched is correct, not a failure. Some are built as time sinks.
People also search for
What is the best order to study CAT quant topics?
Arithmetic first, then algebra, then geometry and number system, with modern maths last. Arithmetic is the largest block and it compounds — percentages underpin profit and loss, interest, mixtures and most data interpretation, so fluency there pre-solves part of several other topics. Modern maths, meaning permutation, combination and probability, is the smallest block and the slowest to make fast, which makes it a poor use of early months. The commonest sequencing error is doing modern maths first because it feels like real CAT maths.
What level of maths is needed for CAT quant?
Roughly class 10, applied at speed. There is no calculus, no matrices or determinants, no advanced trigonometry and no statistics beyond mean, median and mode. What makes CAT quant hard is the time available — around 109 seconds per question — and the fact that questions are written to punish mechanical approaches. A problem that looks like a long calculation usually has a shortcut, and finding it is the actual test rather than the arithmetic itself.
How many QA questions should I attempt in CAT?
Most candidates score best attempting 14 to 18 of the 22 questions at 80 percent accuracy or better. Leaving four untouched is a correct outcome rather than a failure, because some CAT quant questions are deliberately built as time sinks and recognising one is worth more than solving it. Test this range against your own mock data by plotting attempts against final marks across five mocks — the point where marks stop rising is your personal optimum.
Why do engineers struggle with CAT quant?
Not because of knowledge — engineers usually have far more mathematics than CAT requires. The problem is habits. Engineering trains you to solve rigorously and completely, reach for the general method, distrust approximation, and finish what you started. All four cost marks under a 109-second budget, where the winning approach is to recognise the question type, apply a shortcut, estimate when options are far apart, and abandon anything that is solvable but not solvable in time. Adjusting those habits early usually produces fast improvement.
Which arithmetic values should I memorise for CAT?
Multiplication tables to 20, squares to 30, cubes to 15, percentage-to-fraction conversions from a half down to a sixteenth, powers of 2 and 3, common Pythagorean triples, and primes up to 100. This feels like school work and it is the highest-return week of preparation available. Finding 37.5 percent of 4,896 takes about 25 seconds by long multiplication and about 6 seconds if you know 37.5 percent is three-eighths. Across a whole section that difference is two extra questions attempted.
Should I attempt CAT quant questions in order?
No. Use three passes. In the first twenty minutes solve everything you recognise immediately, skipping anything where you cannot see how to start within about fifteen seconds. In the next thirteen minutes return to skipped questions where an approach has since occurred to you, prioritising TITA questions since they carry no penalty. Use the final five minutes for educated guesses on MCQs where you have eliminated two options, and to fill every remaining TITA. Stop attempting with two minutes left.
How long should I spend on one CAT quant question?
The section average is about 109 seconds, but averages mislead — you should be finishing recognised questions in under a minute so that harder ones can take longer. The rule that matters more is the start rule: if you cannot see the opening move within fifteen seconds, skip it on the first pass. Questions where the approach is not immediately visible are exactly the ones that consume four minutes and return nothing, and skipping them is how you protect the time the rest of the section needs.
Is arithmetic or algebra more important for CAT?
Arithmetic, clearly, and by a larger margin than most candidates expect. It carries the largest share of QA questions, it compounds into algebra and geometry problems that involve ratios or percentages, and it feeds directly into DILR caselets where the reasoning is structural but the calculation is arithmetic. A candidate strong in arithmetic and weak in modern maths reliably outperforms the reverse, because modern maths is a small block that also happens to be the hardest to make quick.
Can I skip modern maths in CAT preparation?
Not entirely, but it should be last and it should get proportionate time. Permutation, combination, probability and set theory make up the smallest block in QA, and they are the slowest topics to become fast at because each question tends to need fresh reasoning rather than pattern recognition. Study them after arithmetic, algebra, geometry and number system are solid. Skipping them completely costs you a small number of questions; doing them first costs you fluency in the areas that carry far more.
How do I improve speed in CAT quant?
Three things in order. First, memorise the arithmetic furniture so calculation stops being the bottleneck. Second, practise topic-wise until you recognise question types on sight rather than working out which method applies. Third, build the habit of checking how far apart the options are before deciding how precisely to calculate, because CAT frequently spaces them widely enough that estimation identifies the answer. Speed in CAT quant comes from recognition and estimation, not from calculating faster.
Like what you read? Share with a friend
Senior Content Writer
Co-Founder at PrepGrind, working on education content, product development, and student-focused learning resources. I also write articles and guides for MBA/MBS entrance exams, covering preparation strategies, exam updates, and useful resources for students.
