Method plus 2026 reference points. No 2027 data exists yet

JEE Main 2027 Marks vs Percentile: How NTA Converts Scores

NTA converts marks into a percentile by ranking you inside your own session: percentile = 100 x (candidates with a score at or below yours) / (all candidates in that session). That is why no fixed marks-to-percentile table exists. This page explains the method, shows what 2026 looked like and how to estimate without fooling yourself.

Last verified: 15 November 2026

Key facts at a glance

Formula
100 x (candidates scoring <= you) / (total candidates in that session)
Is it fixed?
No. It changes with each session, shift and year
2026 Session 1 appeared
About 13.05 lakh (reported: 13,04,653)
2026 General JEE Advanced cut-off
93.4123549 percentile (NTA, 20 April 2026)
2027 Session 1
20, 21, 22, 23, 24 and 28 January 2027
Status (15 Nov 2026)
No 2027 marks-percentile data exists; 2026 shown for reference only

Status: what exists and what does not

As of 15 November 2026 no JEE Main 2027 paper has been held, so there is no 2027 marks-vs-percentile data. Everything labelled 2026 below describes the previous cycle and is a reference, not a prediction. Session 1 is on 20, 21, 22, 23, 24 and 28 January 2027.

How NTA converts raw marks into a percentile

The NTA score in JEE Main is a percentile, not a percentage. For a given session, NTA lists all candidates by their score and gives each candidate the share of the session's candidates who scored equal to or less than them. In formula form: percentile = 100 x (number of candidates with score less than or equal to yours) / (total number of candidates in that session).

A candidate at the 99th percentile has therefore beaten or tied with 99 per cent of the people who appeared in that session. With about 13.05 lakh candidates in Session 1 of 2026, each percentile point near the top stands for roughly 13,000 candidates, so a single mark can shift your percentile by a visible amount in the dense middle of the distribution and by a tiny amount at the very top.

When a session is spread over several shifts, the raw scores of different shifts are not directly comparable, because the papers differ. Per Appendix IV, the highest raw score in every shift becomes 100 percentile, the lowest gets a small value that depends on the number of candidates in that shift, and the shift percentiles are merged and called the NTA scores, which then drive the merit list. Our page on the normalization process explains this in more detail.

The bulletin works an example for a shift of 41,326 candidates: a candidate with 121 raw marks and 37,244 candidates at or below that score gets (37,244 / 41,326) x 100 = 90.1224411 percentile, while 41 raw marks with 20,851 at or below gives 50.4549194. Percentiles are carried to 7 decimal places to avoid bunching and reduce ties.

Eligibility cut-offs follow the same logic: if the cut-off marks correspond to different percentiles in different shifts, the lowest percentile becomes the cut-off for that category in all shifts (bulletin example: 78 in one shift and 79 in another makes 78 the cut-off). Paper 1 merit ties are broken by NTA score in Mathematics, then Physics, then Chemistry, then lower proportion of incorrect to correct answers (bulletin 2.4).

Why marks to percentile is not fixed

  • Paper difficulty: a harder shift lowers every score, so the marks needed for 99 percentile fall. In 2026, one coaching analysis estimates the 99th percentile mark ranged from about 154 in the toughest Session 1 shift to about 188 in the easiest.
  • Candidate pool: a stronger pool pushes the same marks to a lower percentile. Session 2 attracts different candidates than Session 1 (some skip it, some are repeaters improving a score).
  • Distribution shape: marks are dense between 40 and 100 and thin at the top, so a 10-mark change moves the percentile a lot in the middle and very little at the top.
  • Negative marking: Paper 1 awards +4 for a correct answer and -1 for a wrong one in both the MCQ and numerical-value sections (bulletin 2.4), so the spread depends on how many guesses go wrong.
  • Year to year: attendance, difficulty and the share of repeaters change. The average 99th-percentile mark in one analysis rose from about 170 in 2025 to about 174 in 2026, which shows a drift even when the paper style is the same.

2026 reference points (labelled 2026, not 2027)

The tables below are what the 2026 cycle looked like according to published analyses. They are shift-wise estimates by private organisations, not NTA data. Where two sources overlap they broadly agree but are not identical, which is itself a reminder not to use a single number.

2026 raw score bandApprox. percentile (2026 Session 1)Approx. rank (2026)Source
300 to 291100 to 99.999891 to 15PW result analysis
290 to 28199.999 to 99.99716 to 36PW result analysis
270 to 261about 99.990101 to 160PW result analysis
200 to 191about 99.6 to 99.73,804 to 5,320PW result analysis
160 to 151about 98.7 to 99.013,164 to 17,290PW result analysis
100 to 91about 94.3 to 95.061,652 to 79,298PW result analysis

Approximate figures from a coaching site's 2026 analysis (published after the Session 1 result). They are not published by NTA. PW itself calls them approximate.

99th percentile marks by shift in 2026 (estimates)

Read the spread, not the individual cells. Inside one session the 99th percentile mark moved by more than 30 marks from the hardest to the easiest shift. A candidate who plans for exactly 160 marks and lands in an easier shift could miss the 99th percentile by a visible margin.

Session 1 shift (2026)99th percentile marksSession 2 shift (2026)99th percentile marks
21 Jan, shift 11582 Apr, shift 1175
21 Jan, shift 21742 Apr, shift 2165
22 Jan, shift 11864 Apr, shift 1175
22 Jan, shift 21544 Apr, shift 2174
23 Jan, shift 11665 Apr, shift 1190
23 Jan, shift 21585 Apr, shift 2175
24 Jan, shift 11746 Apr, shift 1160
24 Jan, shift 21786 Apr, shift 2183
28 Jan, shift 1188--
28 Jan, shift 2180--

Competishun's own shift-wise estimates for 2026, not official NTA data. The same page gives roughly 230 marks for the 99.9th percentile and roughly 113 (January) and 124 (April) marks for the 95th percentile on average. Its summary range and its band table differ slightly.

The only official 2026 percentile anchors: JEE Advanced cut-offs

NTA published category-wise cut-off percentiles on 20 April 2026 for JEE Advanced 2026 qualification, in the roughly 2.5 lakh candidate pool. These are percentiles, not marks, and they come from the combined result of both sessions.

Category (2026)Cut-off percentile for JEE Advanced 2026
General93.4123549
Gen-EWS82.4164528
OBC-NCL80.9232583
SC63.9172792
ST52.0174712

2026 values as published by NTA on 20 April 2026. 2027 cut-offs will be published only after both 2027 sessions.

How to estimate your percentile safely after the exam

  1. 1

    Compute your raw marks from the response sheet

    After each session NTA releases response sheets and a provisional answer key. Count +4 for each correct answer and -1 for each wrong answer in both Section A (MCQs) and Section B (numerical-value questions); unanswered questions score 0 (bulletin 2.4). See the Session 1 answer key page for the process.

  2. 2

    Locate your shift and its difficulty

    Read the shift-wise analysis for your date and shift. Note whether your shift was reported as easy, moderate or difficult by several sources, not one.

  3. 3

    Use a band, not a point

    Take the nearest 2026 reference rows and widen them by about 10 to 15 marks on each side. If your shift was easier than the 2026 average, move your expected percentile down; if harder, up.

  4. 4

    Cross-check with two or more sources

    If two independent coaching sites give similar bands for your shift, the estimate is more reliable. If they differ widely, assume the lower percentile for planning.

  5. 5

    Plan on the lower end

    Use the lower end of your band for deciding whether you need Session 2 and for first college lists. Replace the estimate with the real percentile the moment the scorecard is out.

Common mistakes when using marks vs percentile

  • Reading a marks-to-percentile table as an official NTA conversion. NTA does not publish one.
  • Using a 2025 or 2026 table for 2027 and treating it as a target. Difficulty changes every year.
  • Adding or averaging your Session 1 and Session 2 marks. Only the better percentile is used.
  • Ignoring the shift. The same marks in two shifts can differ by several percentile points.
  • Treating the percentile as percentage of marks. A 90 percentile does not mean 90 per cent marks.
  • Quoting a rank from a percentile without knowing the number of candidates. See the percentile to rank page.
  • Calculating marks from a wrong answer key. The provisional key can change after challenges; recheck after the final key.

Worked example with 2026 numbers

Suppose a candidate in 2026 Session 1 estimated their score at about 160. The PW analysis lists 160 to 151 marks at roughly the 98.7 to 99.0 percentile. The Competishun shift table says the 99th percentile mark was about 158 in the 21 January morning shift but about 186 in the 22 January morning shift. So the same 160 marks might be just above 99 in one shift and nearer 98 in another.

The safe reading is therefore a band of about 98 to 99.2, and the candidate should plan with the lower end. If their aim was the General cut-off for JEE Advanced (93.41 in 2026), they were comfortably above it on any reading. If their aim was a high-ranking NIT branch, the extra decimal places of percentile matter and they should use Session 2.

What to watch before using any 2027 number

  • Shift-wise analyses after each January 2027 shift, from at least two sources.
  • The NTA result and scorecard, which is the only exact conversion for your own marks.
  • The Session 2 timetable (2, 3, 4, 5, 6 and 7 April 2027), if you plan to attempt both sessions; its form opens on 5 January 2027.
  • The pattern in bulletin 2.4: 75 questions and 300 marks for Paper 1, with 25 questions per subject.
To go from percentile to expected rank, see the JEE Main 2027 percentile to rank page. For a college-level view, use our college predictor once your real percentile is out.

Sources

JEE Main 2027: complete guide

Dates, eligibility, pattern, syllabus, result and counselling in one place.

Open guide

Frequently Asked Questions

Answers on marks vs percentile, from the CSAB 2026 brochure

1
How does NTA convert JEE Main marks into a percentile?
Appendix IV of the 2027 bulletin gives the formula: percentile = 100 x (number of candidates appeared in the shift with raw score equal to or less than the candidate) / (total number of candidates who appeared in the shift). It is computed shift by shift on the total raw score, to 7 decimal places, and the percentile of the total is not an aggregate or average of the subject percentiles. The percentile is not the same as the percentage of marks.
2
Is there a fixed marks-to-percentile table for JEE Main 2027?
No. The percentile depends on how every other candidate performs in the same session, so the same marks give different percentiles in different years and sessions. A table can only describe what happened in a past cycle. No 2027 table can exist before the January 2027 exam is held.
3
How many marks are needed for 99 percentile in JEE Main?
In 2026, a coaching-institute analysis (Competishun) estimated the 99th percentile at roughly 154 to 188 marks across the ten Session 1 shifts and 160 to 190 across the eight Session 2 shifts. These are shift-wise estimates by a private site, not NTA figures. Treat them as a range that shows how much the answer moves with paper difficulty.
4
Why does the same score give different percentiles in different shifts?
Because each shift is a different paper with a different set of candidates. A harder shift pushes everyone's marks down, so a given score sits higher in that shift's order. The bulletin says the normalization is used so that candidates are neither benefitted nor disadvantaged by the difficulty of their shift, and the percentile (the NTA score), not the raw marks, is what is merged and compared.
5
Does a higher percentile always mean more marks?
Within one shift, yes. Across shifts, not necessarily. A candidate with fewer marks in a harder shift can end up with a higher percentile than a candidate with more marks in an easier shift.
6
Which percentile counts if I appear in both sessions?
Bulletin 11.3.1 (c): the best of the two NTA Scores in Total (and not in individual subject) is considered for candidates who appeared in both sessions. There is no averaging, and marks from the two sessions are never added. The Common Rank List is compiled only after Session 2 (Note 11.3.2).
7
Can I trust the marks-vs-percentile tables on coaching sites?
Use them as rough guides only. Tables built after the result are estimates from sampled student data or from shift-wise reports, and different sites give different numbers for the same percentile. NTA's scorecard shows the NTA score, so the only exact conversion is the one on your own scorecard. The bulletin provides no marks-to-percentile table.
8
Is the 2026 table valid for JEE Main 2027?
It is a reference for how the scale behaved, not a prediction. If the 2027 paper is easier, the same percentile will need more marks; if harder, fewer. Plan with a margin rather than a single target mark.
9
What was the JEE Advanced qualifying percentile in 2026?
NTA published category cut-offs on 20 April 2026 for JEE Advanced 2026: General 93.4123549, Gen-EWS 82.4164528, OBC-NCL 80.9232583, SC 63.9172792 and ST 52.0174712 percentile. These are 2026 values. The 2027 cut-offs will be known only after the 2027 sessions.
10
Where do I see my percentile after the exam?
On the NTA scorecard at jeemain.nta.nic.in, after the result is declared; no score or rank card is dispatched, so download it from the site (Note 11.3.2). A copy of the final NTA score is also sent to your registered email. The scorecard carries the name as in your Class 10 certificate. The response sheet released before the result lets you estimate raw marks, but not the percentile.
College Predictor

Know Your College
Before Counselling

Predict your college using JoSAA 2026 cutoff data across IITs, NITs, IIITs and GFTIs.

135+
Institutes
3000+
Programmes
2026
Cutoff Data
JEE Main
JEE MainNITs · IIITs · GFTIs
JEE Advanced
JEE AdvancedAll 23 IITs
DASA / NRI
DASA / NRIForeign nationals
Predict My College

Source

The percentile formula and normalization steps are from Appendix IV of NTA's JEE Main 2027 Information Bulletin; the 2026 category cut-offs were published by NTA. Marks reference points are from coaching-institute analyses (Competishun and PW), not from NTA. Last checked 15 November 2026.