NCET Cut-Off, Previous Papers and Answer Keys: What Actually Exists

ProSyllabus is an independent study site. We are not affiliated with, or endorsed by, the National Testing Agency, the National Council for Teacher Education or any institution named here. Every rule below is quoted or summarised from their published documents, and the documents control if anything here differs.

Nothing for NCET 2027 has been announced: when we checked the NTA portal on 6 October 2026 its newest public notice was the result notice of 4 June 2026, and no 2027 bulletin, fee or date was posted. Every 2027 line on this page is a projection and says so.

What NTA publishes, and what it does not

How the NTA score is built

Because the test runs in several shifts and subjects, NTA does not give raw marks. The 2026 bulletin's Appendix VI explains the method. Within each shift and for each subject, every candidate's raw marks become a percentile: count the candidates in that shift whose raw score is less than or equal to yours (m), divide by the number who appeared in the shift (N) and multiply by 100. Those percentiles are then pulled back to a marks scale to normalize the shifts, using the equi-percentile method, so that a harder shift does not penalise its candidates. The appendix says the method follows a committee headed by a senior professor from the Indian Statistical Institute, Delhi, with members from IIT Delhi and the University of Delhi, and cites the 2020 Current Science paper by Bhatt and others. The bulletin calls the result the NTA score: "Percentile Score and Normalized Score".

A hundredth percentile is not rare: NTA's 2025 table

NTA's press release of 16 June 2025 lists, for every paper, how many candidates registered, how many appeared and how many sat at the 100th percentile. The numbers make the point that the top percentile is a tie, not a single winner. Our division of the 100-percentile count by the number who appeared is in the last column; NTA does not explain the counts, but because a percentile is calculated inside each shift and ties at the top raw score all receive 100, papers with few candidates or an easy shift can show many.

Answer keys: what the PDFs contain

NTA's final answer key is a list, by subject and shift, of Question Id and the key. There is no question text and no explanation. The 2026 key has 141 pages. By our parse it lists 3,178 questions in 76 subject or section blocks over three shifts: 17 April first shift, 17 April second shift and 18 April first shift, which matches the three shifts in NTA's city notice. Each shift of the General Test has 28 questions, and the keys for the three shifts have different Question IDs, so each shift had its own paper. The 2025 key has 73 pages and, by our parse, 3,654 question rows.

The special keys are the interesting ones. A key such as "1,3" means two options were accepted as correct; under the marking scheme anyone who marked either gets +4. "Drop" means the question was removed and, under the scheme, every candidate who attempted it gets +4. NTA does not say which of these came from challenges. Our count of the 2026 file found these seven entries; the Question Id is as printed.

Of the 2025 challenges, 191 on 124 questions across 19 subjects changed the keys for three questions, which is 2.4 per cent of the questions challenged (our division). The 2026 numbers are not in the public notice, so we do not guess them. The practical lesson is the one in the notice: a challenge costs Rs 200 per question, is non-refundable, must come with justification or evidence, and has to be made inside a three-day window, so challenge only where you have a source.

There are no public previous papers: what to do instead

The answer positions in the keys: a pattern we measured

While counting the keys we also counted which option number is correct. Across the 3,171 single-option keys of 2026 and the 3,648 of 2025, the middle options are correct more often than the outer ones. In both years options 2 and 3 together hold about 56 per cent of the keys, against 50 per cent if the four options were equally likely. A chi-square test of equal shares gives 51.1 for 2026 and 53.2 for 2025 with three degrees of freedom; values above about 11.3 would happen by chance less than one time in a hundred.

Why this is not a strategy. First, the file prints "Answer Key" or "Correct Option ID" as a number from 1 to 4; it does not say whether that number is the option's position on your screen, so we cannot be sure it maps to what a candidate saw. Second, a pattern in two years of keys is a fact about two sets of question writers; it may not hold next year, and it is weaker inside any one paper (the 84 General Test keys of 2026 do not show it, with a chi-square of 2.2). Third, even if you guessed only 2 or 3 and the true rate was 28 per cent, your expected score on a blind guess would be 0.28 x 4 minus 0.72 x 1 = +0.40 a question, against +0.25 for a random guess; real preparation raises your accuracy by far more. We include the count because it is checkable, and so you can see that a key is data, not a coin.

Where closing scores can come from, and the competition per seat

If you want a number to aim at, the sources are institutional: a merit list or a closing score published after counselling. For NCET 2026 we hold none, because the three brochures we read describe the rules, not past results. Before choosing where to apply, ask the institution for last year's merit list and the seat matrix by category, and compare those against your own percentile and tie-break standing.

A national figure can only give a very rough sense of competition. The table divides the number of candidates who appeared by the seat totals printed in each document; it is our arithmetic and a poor measure, because seats are spread over about 280 institutions and candidates can apply to more than one.

The ratio fell sharply because the institution list and the seat totals grew faster than the number of candidates: registrations rose about 35 per cent from 2025 to 2026, while the seat totals above rose between 178 and 339 per cent. The figure says nothing about any one college. Any single institution can be far tighter or looser than the national figure.