1,002 CAT questions tagged across all sections (2021–2025). Here's where the marks actually come from.
A cross-section analysis of every question from CAT 2021–2025. RC Inference is the single most tested skill in the paper. Tables-based DI is second. Most prep plans have the emphasis backwards.
The full picture
1,002 questions. 15 slots. CAT 2021–2025. Every question tagged by section, topic, and subtopic. Here's what the data shows when you look at the paper as a whole.
Section breakdown
| Section | Questions | % of paper | Time (min) |
|---|---|---|---|
| VARC | 360 | 36% | 40 |
| Quant | 330 | 33% | 40 |
| DILR | 312 | 31% | 40 |
Questions per section (1,002 total)
Each section gets the same time despite unequal question counts. DILR has fewer questions per minute than VARC — each DILR question requires more time on average.
The 10 highest-frequency skills in the entire paper
Ranked by number of questions across 2021–2025:
| Rank | Skill | Section | Questions | % of paper |
|---|---|---|---|---|
| 1 | RC Inference | VARC | 117 | 11.7% |
| 2 | DI Tables | DILR | 72 | 7.2% |
| 3 | QA Arithmetic (all) | Quant | 115 | 11.5% |
| 4 | QA Algebra (all) | Quant | 101 | 10.1% |
| 5 | RC Detail / Factual | VARC | 44 | 4.4% |
| 6 | DILR Arrangements | DILR | 51 | 5.1% |
| 7 | RC Main Idea | VARC | 28 | 2.8% |
| 8 | QA Geometry (all) | Quant | 50 | 5.0% |
| 9 | Para Summary | VARC | 39 | 3.9% |
| 10 | DI Matrix / Grid | DILR | 29 | 2.9% |
Top 10 skills by question count (2021–2025)
RC Inference alone is 11.7% of the entire paper. One skill, correctly trained, impacts more than one-tenth of your total score.
VARC: 360 questions
| Topic | Questions | % of VARC |
|---|---|---|
| Reading Comprehension | 240 | 67% |
| — Inference | 117 | 49% of RC |
| — Detail / Factual | 44 | 18% of RC |
| — Main Idea | 28 | 12% of RC |
| — Tone & Attitude | 12 | 5% of RC |
| Para Summary | 39 | 11% |
| Para Jumble | 35 | 10% |
| Sentence Insertion | 22 | 6% |
| Odd Sentence Out | 18 | 5% |
| Para Completion | 6 | 2% |
Quant: 330 questions
| Topic | Questions | % of QA |
|---|---|---|
| Arithmetic | 115 | 35% |
| Algebra | 101 | 31% |
| Geometry | 50 | 15% |
| Number System | 41 | 12% |
| Modern Math | 23 | 7% |
DILR: 312 questions
| Topic | Questions | % of DILR |
|---|---|---|
| Data Interpretation | 187 | 60% |
| — Tables | 72 | 23% of DILR |
| — Matrix / Grid | 29 | 9% of DILR |
| — Networks | 25 | 8% of DILR |
| — Bar Charts | 21 | 7% of DILR |
| Arrangements | 51 | 16% |
| Logical Reasoning | 50 | 16% |
| Scheduling / Networks / Other | 24 | 8% |
Three mistakes this data reveals
1. Treating RC as one undifferentiated skill.
RC Inference (117 questions) and RC Detail (44 questions) require different approaches — Inference tests logical reasoning from the passage, Detail tests close reading. Practising "RC" as a whole skill without separating question types is the most common VARC mistake.
2. Under-preparing Arithmetic in QA.
Time & Work (14 questions) and Averages (13 questions) are bigger than most coaching plans acknowledge. Ratios & Proportions gets significant attention but is only 8 questions. The coaching emphasis and the actual paper are misaligned.
3. Assuming DILR = Arrangements.
DI (Tables, Matrix, Charts) is 60% of DILR. If you've spent 80% of your DILR time on seating and linear arrangements, your allocation is backwards. Tables alone (72 questions) exceed all Arrangements combined (51 questions).
How to use this data
These numbers are from the actual PYQ database — you can practice by any topic, section, or question type directly:
- Filter by topic: Go to Past Questions and filter to any subtopic to drill specifically
- Full-length practice: 60 mock tests in CAT 2025 format
- Topic by topic: Study material organised by section and topic
Put this into practice
Solve CAT General PYQs — filtered by topic, with full solutions.