NEET PG 2026: the result is out, and 262 is the bar

24 September 2026

Result declared

641 / 720

AIR 1

614 – 641

Top 20 scored

262 / 720

General / EWS cutoff

1,33,655

Cleared the General bar

From 1 October 2026

Scorecards

NBEMS declared the result on 24 September 2026, about a week ahead of the 30 September commitment, and with the category cutoffs attached rather than following separately. For the October counselling schedule and current notices, see the 2026 counselling updates. Everything below is the paper that produced these numbers, and what they mean for a preference list.

The reduction was about pacing, not syllabus. NMC chairperson Dr Abhijat Sheth described the intent as giving candidates more room per question: rather than extending a 3½-hour paper, the committee removed 20 questions, which buys back roughly seven seconds on each of the ones that remain.

The sectional lock is the change candidates felt most. Each group of 36 questions gets 42 minutes; when they are up, that group closes and the next opens. There is no banking time from an easy section and no returning to a question you meant to revisit. Candidates reported groups one and two as the most time-consuming, which is exactly where the lock bites hardest.

Twenty fewer questions means the same number of candidates spread over a narrower range of possible scores, so more of them land on exactly the same total. NBEMS set out the order in which those ties are broken:

1. Fewer incorrect answers overall.
2. Fewer incorrect answers in Section A.
3. Fewer incorrect answers in Sections B, C, D and E, in that order.
4. Higher aggregate marks in the MBBS professional examinations.
5. The older candidate.

The second criterion drew objections: in a computer-based test where the question sequence varies between candidates, aspirants questioned whether Section A performance is a fair thing to compare. The NMC’s response was that a computer-based format has “many objective tools” available to keep the process fair, and that the exam would be run “in a very safe, secure and transparent manner”.

Moderate, with a large minority saying difficult. Across the published candidate surveys the split was consistent, and the reasons given were about form rather than syllabus — long stems, close options, and application over recall.

On pacing, the extra seven seconds a question seem to have landed: 67.7% of candidates attempted 151–180 of the 180 questions and the same share called the duration adequate, leaving about a third who did not. Roughly 15–20 questions were widely agreed to be genuinely hard.

Two things candidates expected and did not get: video-based questions, which despite the speculation did not appear at all, and image-based questions, which made up only about one in ten. Medicine and Biochemistry were most often named as the hardest subjects — Medicine for the length of its clinical scenarios, Biochemistry for conceptual and case-based items. Gynaecology, Pharmacology, Surgery and Ophthalmology produced the most memorable questions, with recurring topics including vitamin B deficiencies, arthritis, TORCH infections, CMV and sexually transmitted diseases.

The consistent thread is that the paper rewarded clinical reasoning over recall. Estimates of how many questions overlapped with previous years ranged from five to seventy-plus depending on who was counting, so direct repeats were not a reliable strategy — but previous-year papers remained useful for the pattern of reasoning being asked for.

30 August 2026

Exam date

9:00 am – 12:30 pm, single shift

Session

2,73,083

Registered

2,65,960

Appeared

1,111

Centres

338

Cities

A power failure at the Sitapura centre in Jaipur disrupted roughly 2,445 candidates, who were scheduled for a re-examination on 5 September 2026. Their scores enter the same merit list, so the final rank list is not settled until that sitting is marked.

NBEMS publishes no marks-to-rank table, but for 2026 it published something better: the full result, every candidate’s score against their All India Rank. The anchors below are read off that document rather than projected onto it, so this is the real distribution and not a coaching estimate. It still answers as a range, because one mark is shared by hundreds of candidates — that span is the tie, not our uncertainty.

Each anchor is the deepest rank that mark reached in the published result, so the bands butt up against each other with nothing falling between them. Read a row as “this score puts you in this band”:

Three checks run on this table in the test suite, so a mistyped anchor fails the build rather than reaching the page. It has to be monotonic and bounded by the 2,65,960 who appeared; it has to reproduce the declared qualifying marks, which were counted separately and which it currently matches to within 0.05%; and converting a 720-mark score to its 800-mark equivalent, reading the 2025 curve and scaling by the growth in candidates has to land in the same neighbourhood — the guard that would catch a curve applied across the scale change.

On 3 September 2026 — three days after the paper and three weeks before the result — this page published a projected version of that table, quoting every answer as a range of ±45%. The projection is still in the data file rather than quietly deleted, and the table below replays it against what NBEMS actually published. A predictor that only ever shows you its corrected answers is not one you can calibrate your trust in.

7%

median error

−6%

average bias, too kind

22/23

inside the quoted range

56%

worst single miss

The short version: it held up where it decides anything, and drifted where the stakes are lower. Through the 400–530 band — the range that actually settles who gets an MD or MS seat — every projection landed within 6.5% of the truth, and at 500 marks it was out by 371 ranks in 12,111. Below 250 marks it ran consistently about 10% generous, telling a candidate their rank was better than it was; that is the wrong direction to be wrong in, but at those depths it changes nothing about which seats are reachable.

The one real failure is at the very top. The projection put 600 marks at rank 50; the result puts it at 114, so it was more than twice as optimistic as it should have been for the few hundred candidates competing for the most contested seats in the country — exactly the people who can least afford a flattering number. Too few candidates score that high for a pre-result estimate to have any resolution there, which is an explanation rather than an excuse, and it is the single anchor of twenty-three that fell outside the ±45% we quoted.

We also got the qualifying cutoff wrong, and not narrowly. This page argued the General bar would land somewhere in the low 210s to high 240s once the harder paper pushed the distribution down. It came in at 262 — above the top of our bracket. The reasoning was sound and the premise was not: the distribution did not shift down nearly as much as a paper rated moderate-to-difficult by 95% of candidates suggested it would.

NBEMS does not set the qualifying bar in marks. It sets it as a percentile of the candidates who appeared, and the mark that percentile lands on is whatever the distribution says once every paper is scored. The percentiles are unchanged for 2026; the marks they landed on are new:

Read across the row rather than down the mark column. The General bar fell from 276 to 262, which looks like an easier year and is not: as a share of the marks available it went from 34.5% to 36.4%, so relative to the paper the bar went up. The two numbers are on different scales and the only thing they genuinely share is the percentile.

The last column is the reason the marks-vs-rank table above can be checked at all. The count of candidates at or above a bar is the deepest rank on it, so each declared cutoff is one point on the score-to-rank curve stated twice — once as a mark, once as a rank — and the two are counted independently of each other. Our curve reproduces all three to within 0.05%, and the test suite fails the build if that drifts past half a percent.

Worth knowing: a number of sites are still quoting NEET PG 2026 cutoffs and marks-vs-rank figures out of 800, and some describe the 2026 paper as 200 questions in one breath and 720 marks in the next. If a figure you are reading does not say which scale it is on, it is worth checking before you plan around it.

1. 5 September 2026

   Re-examination for the roughly 2,445 candidates at the Jaipur (Sitapura) centre whose session was disrupted by a power failure. 2,416 sat it, bringing the total who appeared to 2,65,960.

2. 24 September 2026

   Result and category-wise qualifying cutoff declared by NBEMS. AIR 1 scored 641 out of 720 and the top twenty ranks fell between 614 and 641.

3. 1 October 2026

   Individual scorecards on the NBEMS site, carrying your marks out of 720, your All India Rank and your percentile, together with the final answer key. The scorecard is valid for six months.

4. Next

   MCC publishes the seat matrix and opens registration for All India Quota, deemed and central university, and DNB counselling. Dates are not out yet; the rank on your scorecard is the number that whole process runs on.

The scale change matters more for planning than the difficulty does. Counselling runs on rank, not marks — MCC never sees your score — so the pattern change does not alter a single closing rank. What it alters is the translation on the way in: a mark that meant one rank last year means a materially better one this year, and a predictor still holding an 800-mark curve will quietly tell you to aim lower than you should.

Our predictor uses a separate 720-mark curve for 2026 and labels which scale it is on. Give it your marks now, and your real All India Rank when it arrives — the rank is always the sharper input, because it is the number counselling itself uses.

The result itself — every rank in the marks-vs-rank table, the qualifying marks and the counts that cleared them — is read from the NBEMS published result. The difficulty and subject-weightage numbers and the candidate quotes are as reported by the outlets below; those survey percentages are self-reported, so treat them as the shape of the response rather than a measurement. The interpolation between anchors, the scale-change comparisons and the consistency checks on both are ours.

• Careers360 — NEET PG 2026 exam analysis, answer key and cutoff live coverage
• Careers360 — NEET PG 2026 exam difficulty level
• PrepLadder — NEET PG 2026 detailed exam analysis (candidate survey)
• Careers360 — Revised pattern and tie-break rules; NMC clarification
• Reflex Prep — NEET PG 2026 question paper analysis and difficulty
• India TV — NEET PG paper analysis 2026, section-wise difficulty
• PW — NEET PG 2026 exam pattern revised, questions cut from 200 to 180
• NBEMS — Result of NEET-PG 2026, the published result (5,462 pp., 24 September 2026)
• Careers360 — NEET PG 2026 cutoff out: category-wise qualifying marks and percentile