In one sentence
JEE Main normalization means NTA ranks you only against candidates in your own shift and converts that position into a percentile, so a tough paper does not hurt you and an easy paper does not help you. Your raw marks never change.
Why JEE Main needs normalization
Between 10 and 13 lakh candidates write JEE Main in each session, across eight to ten shifts with different question papers. No two papers can be equally difficult. If ranks were based on raw marks, a candidate who drew a tough paper would lose out to someone of equal ability who drew an easy one. Normalization removes that luck by comparing each candidate only with others who saw the same paper.
The formula NTA uses
For every shift, NTA calculates each candidate’s percentile as:
Percentile = 100 × (candidates in your shift with raw score ≤ yours) ÷ (total candidates in your shift)
- It is calculated separately for each shift, then all shifts are merged into one merit list.
- It is reported to seven decimal places (for example 98.7654321) to reduce ties.
- The highest scorer in every shift gets 100 percentile, whatever their raw marks.
- If you take both sessions, the better of your two percentiles is used for your rank.
Worked example: a small shift
Imagine a shift with just 10 candidates who scored: 210, 185, 160, 160, 140, 120, 100, 90, 75 and 60.
| Candidate’s marks | Candidates with marks ≤ this | Percentile |
|---|---|---|
| 210 (topper) | 10 | 100 × 10 ÷ 10 = 100 |
| 185 | 9 | 100 × 9 ÷ 10 = 90 |
| 160 (two candidates) | 8 | 100 × 8 ÷ 10 = 80 (both) |
| 140 | 6 | 100 × 6 ÷ 10 = 60 |
| 60 | 1 | 100 × 1 ÷ 10 = 10 |
Two things to notice. Candidates with equal marks get an equal percentile. And the percentile depends only on how others in the same shift scored, not on the paper’s maximum marks.
Real example from JEE Main 2026
In 2026, 99 percentile needed about 154 marks on 22 January (evening shift) but about 190 marks on 5 April (morning shift). A candidate scoring 170 in the January evening shift was above 99 percentile, while the same 170 in the April morning shift was below it, at about 98.3 percentile. Normalization is doing its job here: the January paper was harder, so fewer marks were needed for the same standing.
See the marks needed for every percentile in every difficulty band in the JEE Main marks vs percentile table.
How percentile becomes your rank
After both sessions, NTA takes each candidate’s best percentile and ranks everyone on one list. Your approximate rank is (100 − percentile) × total candidates ÷ 100. With about 15.5 lakh candidates in 2026, every 0.1 percentile near the top was worth about 1,550 ranks. That is why a few marks can move your rank by thousands. NTA applies its published tie-break rules to candidates with identical percentiles. Try the rank predictor to see your own range.
Common myths
- “Normalization cuts my marks.” No. Your raw score is never changed. Normalization only converts your position in the shift into a percentile.
- “An easy shift is bad luck.” Not really. Everyone in that shift had the same easy paper, so you need more marks but you are compared with the same competitors.
- “Percentile is the percentage of marks I scored.” No. It is the percentage of candidates at or below your score.
- “JEE Advanced is normalized too.” No. JEE Advanced is a single exam with the same papers for everyone, so ranks come directly from marks.
Frequently asked questions
Does NTA publish the normalization formula?
Yes. The percentile formula is published in the JEE Main information bulletin. NTA does not publish shift-wise score distributions, which is why marks vs percentile tables are compiled from result analyses.
Can normalization increase my percentile?
Your percentile is fixed by your position in your shift. If your shift was tough, the same marks give a higher percentile than they would in an easy shift.
Which session’s percentile is used for my rank?
The better of your two session percentiles.
Why do two people with the same percentile get different ranks?
NTA breaks ties using its published tie-break criteria. Check the current information bulletin for the exact order.