Attempts needed for each CAT percentile
The table shows total attempts out of 68 at three accuracy levels. It uses the top of the CAT 2025 score range for each percentile, so the numbers lean to the safe side.
| Percentile | Marks (CAT 2025) | At 75% accuracy | At 85% accuracy | At 95% accuracy |
|---|---|---|---|---|
| 99.5 | 93 to 103 | 49 | 44 | 38 |
| 99 | 81 to 90 | 45 | 38 | 32 |
| 98 | 74 to 81 | 39 | 34 | 31 |
| 95 | 60 to 70 | 35 | 30 | 26 |
| 90 | 50 to 53 | 26 | 23 | 19 |
| 85 | 41 to 45 | 22 | 20 | 16 |
| 80 | 37 to 41 | 20 | 19 | 16 |
Section by section at 85 percent accuracy
| Percentile | VARC (24) | DILR (22) | QA (22) | Total |
|---|---|---|---|---|
| 99.5 | 19 | 13 | 12 | 44 |
| 99 | 16 | 12 | 10 | 38 |
| 98 | 15 | 10 | 9 | 34 |
| 95 | 13 | 9 | 8 | 30 |
| 90 | 10 | 7 | 6 | 23 |
| 85 | 8 | 6 | 6 | 20 |
| 80 | 8 | 6 | 5 | 19 |
How the planner works
Score = (3 × correct answers) − (wrong MCQ answers). A wrong TITA answer costs nothing. The planner finds the fewest attempts that reach your target marks at your accuracy. It spreads wrong answers between MCQ and TITA questions in the same mix as the paper, so a wrong answer costs a little under one mark on average.
There is no fixed number of attempts for any percentile. The marks you need depend on the paper, and the marks you get depend on your accuracy. The planner uses the score range from the last CAT, so the answer moves when either one changes.
Accuracy matters more than volume. At 90 percent accuracy you need several fewer attempts than at 75 percent for the same score, because every wrong MCQ takes a mark away as well as wasting two minutes.
The section split follows how the marks were shared at each percentile in the last CAT: about 44 percent from VARC, 30 from DILR and 26 from QA. If one section is your strength, move a few attempts there, but keep each section above the IIM sectional cutoffs.
Sources
- Cracku: CAT score vs percentile
- MBAUniverse: IMS, Cracku, CL and TIME compared
- Career Launcher: CAT marks vs percentile
Figures last checked on 11 Oct 2026. They are estimates and change with every paper. Your official scorecard is the only exact figure.
