CAT score vs percentile

Three numbers, three jobs

Most of the confusion around CAT results comes from mixing up three different numbers:

IIMs shortlist on the scorecard. Your raw score is a planning tool for December and nothing more.

Data as of: October 2026. Marking rules from the CAT 2026 FAQs.

Why normalisation exists

CAT 2026 runs in three sessions on 29 November 2026, each with a different paper (CAT 2026 Information Bulletin). Papers set by the same committee still differ in difficulty. Without adjustment, a candidate whose session got the harder DILR sets would lose shortlists to someone of equal skill in an easier session.

The IIMs normalise section-wise and overall raw scores across sessions to produce scaled scores, then convert the scaled scores into percentiles. The working assumption behind this kind of normalisation is that each session draws a similar mix of ability, because sessions are allotted by city and centre availability, so differences in score distributions mostly reflect paper difficulty.

The IIMs do not publish the full formula each year. What you can rely on:

• Scaling happens per section as well as overall.
• Your overall scaled score is built from your sectional scaled scores.
• Percentiles are calculated on scaled scores across all test-takers, not just your session.

What a percentile actually means

A 95 overall percentile means about 95% of test-takers scored below you. It says nothing about the share of questions you got right.

That has two practical consequences:

1. A high percentile can come from a modest raw score. In a hard year, a raw score that looks low can still clear 99. Comparing your raw score with a friend’s from a different year or session tells you nothing.
2. Percentiles bunch up at the top. With well over two lakh test-takers (CAT 2025 had about 2.58 lakh, per the 1 December 2025 media release), each 0.1 percentile near the top covers a couple of hundred candidates. The raw-mark gap between 98 and 99 is often far smaller than the gap between 80 and 90.

Why the same raw score lands differently

Two candidates with an identical overall raw score can get different percentiles because:

• Their sessions differed in difficulty. Scaling lifts scores from the harder session.
• Their section mix differed. Overall scaled score is built from sectional scaled scores, so the same overall raw total split differently across sections can scale differently.
• The years differed. Percentile maps move every cycle with paper difficulty and the strength of the pool.

How to read score-to-percentile tables in December

Coaching institutes publish predicted tables within days of the exam. They are built from self-reported scores of their own students and past-year patterns. Treat them as follows:

• Use the band, not the point. A predictor saying “96 to 98” is more honest than one saying “97.2”.
• Expect the most error at the extremes and in the section where sessions differed most that year.
• Never compare across years. A raw score that gave a given percentile in CAT 2025 can give a different one in CAT 2026.

Reading your scorecard

When results arrive (likely in the first week of January 2027, per the bulletin), the scorecard shows a scaled score and a percentile for each of the three sections (VARC, DILR, QA) and overall. Download it the day it opens and save a copy. The CAT 2026 score is valid until 31 December 2027.

Then read it the way an IIM shortlisting committee does:

1. Sectional percentiles first. Most IIMs publish sectional minimums. Missing one by a fraction usually ends your consideration at that campus, whatever the overall.
2. Overall percentile second. This decides which campuses are realistic.
3. Scaled score where the formula asks for it. Some institute formulas convert CAT into points using scaled scores or section-wise components instead of overall percentile. Read each campus’s policy.

See IIM selection criteria explained for how campuses weight CAT against academics, work experience and diversity, and CAT percentile for IIM calls for qualifying floors versus actual call bands.

What this means for your attempts

Because the test is percentile-ranked, your score depends on how you do relative to everyone else, so section balance matters more than one big section:

• Protect each sectional floor. A balanced profile clears more shortlists than a spiky one with a collapsed section.
• Accuracy beats volume. Each wrong MCQ costs one mark and blocks nothing else. Ten attempts at 90% accuracy beat fourteen at 65%.
• Use non-MCQs freely. No negative marking means an educated guess there is free.

Where to go next

• CAT answer key and response sheet to calculate your raw score
• After the CAT 2026 result for the January to March sequence
• IIM cutoffs explained
• Call predictor

Common questions