BharatNotes In CSAT 2024, the split was roughly Maths 35–36, Reading Comprehension 27, and Reasoning 17–18 questions. Across the last decade, comprehension averages 27–30 questions, maths has surged from ~15 (2014) to 33–36 (recent years), and reasoning has shrunk from 30+ to mid-teens. CSAT 2025 confirmed the new normal.
Year-wise question-count table (2014–2025)
Aggregated from Testbook's CSAT trend analysis, Vision IAS solutions, and Edutap year-on-year breakdowns. Numbers are rounded; minor variance exists across coaching classifications.
| Year | Maths / Numeracy | Reasoning | Reading Comprehension | DI / DS | Difficulty |
|---|---|---|---|---|---|
| 2014 | ~15 | ~33 | ~28 | ~4 | Moderate |
| 2015 | ~15 | ~35 | ~30 | ~0 | Moderate |
| 2016 | ~30 | ~13 | ~31 | ~6 | Moderate |
| 2017 | ~26 | ~17 | ~30 | ~6 | Easy |
| 2018 | ~28 | ~24 | ~26 | ~2 | Easy |
| 2019 | ~36 | ~10 | ~27 | ~6 | Easy |
| 2020 | ~40 | ~15 | ~25 | ~0 | Moderate |
| 2021 | ~34 | ~13 | ~30 | ~3 | Moderate |
| 2022 | ~30 | ~20 | 28 | ~2 | Tough |
| 2023 | ~30 | 14 | ~30 | ~6 | Very tough |
| 2024 | 35–36 | 17–18 | 27 | ~10 (woven) | Moderate |
| 2025 | ~33 | ~18 | ~29 | mixed | Moderate-tough |
The four big shifts
- Maths overtook reasoning as the dominant content head around 2016. From 15 questions in 2014 it climbed to a peak of 40 in 2020 and has settled at 33–36 since. Within maths, Number System has emerged as the heaviest sub-topic — CSAT 2025 alone drew 25 questions from Number System.
- Reasoning shrank from 33+ (2014–15) to mid-teens. Direct standalone reasoning puzzles are rarer; reasoning is increasingly woven into data sufficiency or hybrid questions.
- Comprehension is steady at 27–30 questions but harder. Passage length and inferential depth have climbed sharply since 2022. The 2022 paper had 28 RC questions spread across 15 passages, with Level-3 inferential passages reappearing after a 3-year gap.
- Data Sufficiency quietly entered around 2022 and now sits at 8–10 questions a year — often classified inside maths or reasoning, but distinctly testing logical decision-making over computation.
What this means for your prep mix
A sensible weekly time allocation today is roughly:
| Section | Time share | Why |
|---|---|---|
| Maths / Numeracy | 40% | Largest section, biggest scoring opportunity |
| Reading Comprehension | 30% | Steady weight; daily editorial habit suffices |
| Reasoning | 20% | Smaller weight, but high accuracy possible |
| DI / DS | 10% | Small but decisive when papers are tough |
If you're from a non-maths background, flip the maths/RC ratio — give maths 50% and RC 30% in the early weeks, then rebalance once your maths attempt rate stabilises above 25 correct in a mock.
Worked scenario — "weak maths, strong RC"
Say your honest mock baseline is:
- 22 RC correct (out of 27), 1 wrong
- 8 maths correct (out of 35), 4 wrong
- 8 reasoning correct (out of 18), 2 wrong
- 2 DI correct, 1 wrong
That's 40 correct, 8 wrong, 32 skipped. Net = (40 × 2.5) − (8 × 0.833) = 93.33. Already qualifies comfortably. The lesson: a strong RC base does most of the work even if maths is shaky — but only if you've practised RC enough to hit 22+ correct.
Mentor's takeaway
Do not prepare for the 2018-era "easy CSAT." The new normal is 35 maths + 28 RC + 17 reasoning, with 8–10 DI/DS woven in. Build your test-day strategy around that template.
