UPSC Prelims Cut-off Trends 2016-2026 — What Past Cut-offs Tell You

Every aspirant asks the same question in the weeks before Prelims, and most ask it in the wrong way. The question is "what is the cut-off going to be," and it is the wrong question because nobody — not the toppers, not the analysts, not the commission until the moment it publishes — knows the answer in advance. The cut-off is not a target the examiner sets beforehand; it is an outcome that emerges after the exam, fixed by the difficulty of that year's paper, the number of vacancies, and the performance of the entire pool of candidates. The better question, the one this article is built to answer, is "given a decade of past cut-offs, what score should I be aiming for so that I clear regardless of which kind of paper UPSC hands me." That is a question with a usable answer, and the last ten years of data point to it clearly. With the 2026 Prelims written on 24 May 2026 and its result still awaited as of this writing, and with the 2027 Prelims scheduled for 23 May 2027, this is the right moment to understand what the cut-off history actually teaches.

The Decade in Numbers

The figure that matters is the General-category cut-off for General Studies Paper I, because that paper alone, out of two hundred marks, decides who clears Prelims; the aptitude paper is merely qualifying and contributes nothing to this number. Read across the last ten cycles, the cut-off tells a story of steady decline followed by a sharp recovery. In 2016 the General cut-off stood at a formidable 116 marks, the highest of the decade. The following year, 2017, it eased to 105.34. In both 2018 and 2019 it settled at 98 marks, holding steady across two consecutive cycles. The downward drift continued into 2020, when the cut-off fell to 92.51, and then to 87.54 in 2021. The year 2022 saw a marginal uptick to 88.22, before 2023 produced the most dramatic outlier in recent memory: a cut-off of just 75.41, the lowest of the decade, the product of an unusually difficult and heavily current-affairs-and-reasoning-loaded paper that left even strong candidates uncertain. The pendulum then swung back. In 2024 the cut-off recovered to 87.98, and in 2025 it climbed further to 92.66, returning to the territory last seen at the start of the decade's decline. The 2026 figure, following the 24 May 2026 examination, has not yet been released; until the commission publishes it, any number you see attached to 2026 is an estimate, not a fact.

Laid out this way, the ten-year span from 116 down to 75.41 and back up toward the low nineties looks volatile, and in one sense it is. But the volatility is not random noise. It is the visible signature of two forces — paper difficulty and vacancy count — working on a candidate pool that grows more competent every year. Understanding those forces is what turns a table of numbers into a strategy.

Why the Cut-off Moves

The dominant driver of the cut-off is the difficulty of the paper, and the relationship is inverse: a harder paper produces a lower cut-off, an easier paper a higher one. This is not a paradox once you see the mechanism. The cut-off is essentially the score at which the number of candidates above it matches the number of slots available for the next stage, roughly twelve to thirteen times the vacancies. When a paper is brutal, fewer candidates score high, the whole distribution shifts downward, and the line that admits the required number of candidates falls with it. The 2023 cut-off of 75.41 is the textbook illustration: the paper was widely regarded as one of the toughest in years, attempts fell, accuracy suffered, and the cut-off collapsed accordingly. Conversely, the high cut-offs of 2016 and 2025 reflect papers that a well-prepared candidate could navigate with confidence, lifting the whole distribution and pushing the qualifying line up.

The second driver is the number of vacancies. More vacancies mean more candidates admitted to the Mains, which means the qualifying line is drawn lower in the distribution, nudging the cut-off down; fewer vacancies tighten the screen and push it up. The notified vacancies for the 2026 cycle stand at 933, a figure in the moderate band of recent years, neither the unusually high counts that loosen the screen nor the squeezed counts that tighten it severely. The third, quieter driver is the rising competence of the pool itself. The volume and quality of preparation material available to the average serious aspirant has grown enormously over the decade, and a more capable pool, all else equal, pushes cut-offs upward over the long run. This is part of why the recovery from the 2023 trough has been so brisk: the underlying pool is strong, so as soon as the papers normalised in 2024 and 2025, the cut-offs sprang back toward their earlier highs.

A crucial caveat sits underneath all of this. Cut-offs are not strictly comparable across years as raw measures of how much you needed to "know," because the number of questions, the negative-marking arithmetic, and the proportion of the paper given to reasoning and current affairs all shift from year to year. A cut-off is a relative line, not an absolute standard. This is precisely why chasing a specific historical number is a mistake, and why the only sound use of the data is to derive a safe target that survives the worst plausible paper.

What a Safe Target Score Actually Looks Like

Here is where the decade of data becomes directly useful. If you take the ten General-category cut-offs from 2016 to 2025 and ask what score would have cleared in every single one of those years, the answer is governed by the highest of them: 116 in 2016. But 2016 is a genuine outlier from an older era with a different paper structure, and building your entire strategy around clearing the single hardest historical line would push your target into territory that is, for most candidates, unrealistic and unnecessary. The more useful exercise is to look at the cluster the cut-off has actually occupied in recent cycles. Across the last six years, the General cut-off has ranged from the 2023 trough of 75.41 to the 2025 peak of 92.66, with most years sitting in the upper eighties to low nineties.

The logic of a safe target follows directly. To be comfortable rather than anxious, you want a score that clears not the average year but a hard year — because the year that breaks you will be a hard one, and you do not get to choose your paper. A score in the low-to-mid nineties would have cleared every recent General-category cut-off including the 2025 peak, and would have given you a generous cushion in difficult years. The standard advice that emerges from this, and that we endorse, is to treat roughly ninety to one hundred marks as your safe zone for the General category, aiming squarely for the higher end of that band in your mock performance. Hitting the mid-nineties consistently in honest, full-length mocks means you are positioned to clear whatever the commission sets, easy paper or hard. Aiming merely to "scrape past the expected cut-off" is the trap that the 2023 paper sprang on thousands of candidates: those who had calibrated to a comfortable eighty-five found that an unexpectedly tough paper had pushed their actual score well below what they had assumed was safe.

For candidates in the reserved categories the absolute numbers are lower, since the OBC, EWS, SC, and ST cut-offs sit below the General line, but the principle is identical and the gap is smaller than many assume. The right target for any category is set the same way — by aiming to clear a hard year comfortably rather than an average year narrowly — and the disciplined approach is to add a deliberate buffer above your own category's recent cut-off range rather than treating that range as a ceiling.

The Cut-offs You Should Not Confuse With Each Other

A surprising amount of aspirant anxiety comes from mixing up three different cut-offs that the commission publishes, and clarifying them removes a needless source of fear. The first is the Prelims cut-off, the number this article has been discussing, drawn from General Studies Paper I alone and used only to decide who advances to the Mains. The second is the Mains cut-off, set after the written descriptive papers and the essay, which is a different and far higher bar measured on a different total. The third is the final cut-off, computed after the Mains and the Personality Test together, which determines the final selection list. These three are not comparable to one another, and a candidate who reads that "the cut-off was around one thousand" and panics has confused the final cut-off, measured out of more than two thousand marks across many papers, with the Prelims cut-off measured out of two hundred. When you study cut-off trends for Prelims strategy, hold firmly to the fact that the only number relevant to clearing the screening stage is the General Studies Paper I figure out of two hundred, and treat everything else as belonging to a later, separate conversation.

There is a further subtlety worth grasping. The Prelims cut-off is published only after the entire process concludes, often nearly a year after the exam itself, because the commission releases it alongside the final results. This means that during the crucial months when you are actually preparing, the most recent official cut-off available to you is typically one or two cycles old, and every "expected cut-off" circulating in the interim is somebody's estimate built from a sample of answer keys. This is yet another reason to anchor on the broad historical band rather than on any single freshly-circulating prediction. The band is solid ground; the prediction is sand.

What the Category Lines Add to the Picture

While the General-category cut-off is the headline number, the category-wise lines beneath it carry their own lessons for the candidates they concern. Across the decade the OBC cut-off has tended to sit only modestly below the General line, often within a handful of marks, which surprises aspirants who assume a large cushion exists. The EWS line, introduced more recently, has in several years sat very close to or even brushed against the General figure, a reminder that the reservation a candidate is entitled to does not translate into a comfortable margin and that EWS aspirants must prepare to essentially General-category standards. The SC and ST cut-offs sit lower in absolute terms, but the same hard-year logic applies with equal force: in a difficult cycle every category's line falls together, and the candidate who calibrated to an average-year number in their own category is exposed exactly as a General candidate would be. The disciplined takeaway for every category is identical. Identify your own category's recent cut-off range, add a deliberate buffer on top of its upper end, and aim your mock performance at that buffered target rather than at the bare line, because the bare line is precisely the score that fails you in the year that matters.

How to Use This Without Letting It Use You

The danger of cut-off analysis is that it becomes a source of anxiety rather than direction. A candidate who refreshes cut-off predictions in the weeks before the exam is spending energy on a number they cannot control instead of on the preparation they can. The healthy way to hold this data is to let it set one thing — your target score in mocks — and then to forget about it entirely on exam day. Inside the hall, the cut-off is irrelevant; your only job is to maximise your own correct attempts on the paper in front of you, and obsessing over an unknowable qualifying line only corrodes the calm that good performance requires.

Translate the analysis into mock-test behaviour. Take full-length papers under real conditions and score them honestly with the negative marking applied, and track whether your genuine score — not your "I would have got that if I'd had another minute" score — is landing in the mid-nineties. If it is consistently below the upper eighties, that is a signal to deepen accuracy on the high-frequency static topics and to tighten your attempt strategy, not a signal to attempt more questions recklessly. The candidates who clear in hard years are almost always those whose accuracy is high enough that a difficult paper merely lowers their score from excellent to safe, rather than from adequate to failing. Building that accuracy margin is the entire practical lesson of the cut-off data.

It is also worth internalising what the recovery from 2023 to 2025 means for your planning. The pattern suggests that the commission does not let cut-offs stay depressed; a hard year tends to be followed by a normalisation that pushes the line back up. You should therefore not anchor on the 2023 low as your reference point. Prepare for a paper that produces a cut-off in the low nineties, because that is where the recent normal sits, and let the occasional easy year be a pleasant surplus rather than your baseline assumption.

Translating a Target Into Attempts

The reason a target score is more useful than a predicted cut-off is that a target can be reverse-engineered into a concrete plan for exam day, whereas a prediction can only be worried about. Work the arithmetic backward from the mid-nineties and the path becomes visible. With each correct answer worth two marks and each wrong answer costing two-thirds of a mark, reaching roughly ninety-five marks does not require attempting all hundred questions; it requires getting somewhere in the region of fifty-five to sixty questions correct while keeping your wrong answers tightly controlled. A candidate who attempts seventy questions with the accuracy of a well-prepared aspirant clears comfortably, while a candidate who attempts ninety questions on shaky knowledge can actually finish with a lower net score because the negative marking eats the gains. This is the single most important behavioural consequence of cut-off analysis: the goal is never the largest number of attempts but the largest number of correct attempts, and the difference between those two is decided by your accuracy and your discipline in skipping the questions where you are genuinely guessing between two unknowns.

This is why elimination is itself a scoring skill. On a typical paper a substantial fraction of questions can be brought down from four options to two through partial knowledge and reasoning, and on those a calculated attempt carries favourable odds even with negative marking. The questions to leave untouched are the ones where you cannot eliminate even a single option, because there the expected value of an attempt turns negative. Building, in your mock practice, an honest instinct for which bucket a question falls into — confident answer, two-option elimination, or genuine blank — is what lets a target score in the mid-nineties become a reliable outcome rather than a hope. The cut-off data sets the target; your attempt strategy is how you hit it.

A final word on the 2026 and 2027 cycles specifically. With 933 vacancies notified and the recent cut-offs having normalised back into the low nineties, there is no reason to expect the screening line to behave radically differently from the recent pattern unless the paper itself springs a 2023-style surprise. Prepare as though it might. The candidate who builds the accuracy to clear a low-nineties cut-off comfortably is insured against the easy year and the hard year alike, and that insurance, rather than any forecast, is what the decade of data is really urging you to buy.

If you do one concrete thing tomorrow morning, take a full-length Prelims mock under timed conditions, score it strictly with negative marking, and write your honest General Studies Paper I score at the top of the page next to the number ninety-five. The distance between those two numbers is the most important figure in your preparation right now — far more useful than any predicted cut-off — and closing it is the work that actually decides next year's result.

This piece is part of Ease My Prep's Prelims Analysis series; pair it with our companion studies on subject-wise weightage and the most-repeated Prelims topics to turn a target score into a concrete, topic-by-topic revision plan.