UPSC cut-off marks: what score actually clears each stage

UPSC releases CSE cut-offs only after the final result for that cycle, so any number you read for the upcoming CSE 2027 is indicative. Recent General-category bands: Prelims (GS Paper I out of 200) ~75-100, Mains (out of 1750) ~720-770, Final (out of 2025) ~900-1000. CSAT is separately qualifying at 33% and does not contribute to the Prelims cut-off at all.

Civil Services cut-offs are marks, not percentiles. Prelims is out of 200 (GS Paper I only); Mains out of 1750; Final out of 2025 (Mains + Interview). CSAT is separately qualifying at 33%.

What are the indicative CSE cut-offs by category and stage?

How are cut-offs set across the other UPSC exams?

• CDS / NDA: written cut-offs are set per stage and announced with results; SSB then has its own qualifying standard independent of the written marks.
• CAPF AC: a Paper I cut-off screens for Paper II evaluation; physical tests are pass/fail before the interview.
• ESE / CMS: stage-wise cut-offs by engineering stream / medical, released with the final result.

What do the cut-off bands mean for your prep targets?

• Aim ~110+ in GS-I mocks. A comfortable buffer above the worst-case Prelims cut-off absorbs a tough paper.
• Lock CSAT early. Two or three timed CSAT mocks confirm the 33% floor is safe; then ignore it.
• Mains is where rank is built. Once Prelims is safely clearable, answer-writing practice has the highest ROI.

How the Prelims cut-off is set each year

UPSC does not publish a fixed Prelims cut-off in advance. After GS Paper I is evaluated, the cut-off is fixed to filter the top ~12-13 times the expected vacancy count into Mains. Three variables move the bar.

• Paper difficulty: a tough paper compresses the score distribution, pulling the cut-off down. An easy paper widens the spread and pushes it up. Recent years have seen the General cut-off swing in the band of roughly 75 to 100 out of 200 - a 25-mark window, or about 12-13 questions of difference.
• Vacancy count: UPSC selects a Mains pool of roughly 12-13 times the advertised vacancies. A high-vacancy year (1000+) increases the Mains intake and softens the cut-off; a low-vacancy year tightens it.
• CSAT qualification filter: candidates who don't clear CSAT's 33% are removed from the merit pool first. In a year when CSAT spikes in difficulty, more borderline GS scorers fall out at this gate, which can pull the GS cut-off marginally down (because the surviving pool is smaller).

GS Paper I cut-off band in plain terms

For the General category, an indicative band based on recent years is roughly 75-100 out of 200. The most common comfort target inside mocks is 110+, which builds a 10-15 mark buffer over the typical real-paper cut-off. The same logic, with category relaxations applied, gives you a target zone for any aspirant - though every aspirant should mentally aim at the General band, not a lower category number.

• Why aim for the General cut-off? Service allocation prefers candidates in the merit order. Aspirants who only cross their own category cut-off enter Mains in a weaker position, and they need a Mains performance that compensates.
• Why mocks should be harder than the real paper: tougher mocks pull your conservative attempt-count down, calibrate your negative-marking discipline, and ensure that a real paper feels like a step down in difficulty, not a step up.

Mains and Interview - the cut-offs that actually decide rank

The Prelims cut-off only opens the door. The cut-off that determines final selection is the aggregate of Mains (1750) + Interview (275). The final-list cut-off has, in recent years, sat in the rough band of 900-1000 out of 2025 for General candidates - varying with paper difficulty, evaluation rigour and vacancy count.

Numbers above are indicative bands compiled from publicly released UPSC cut-offs over recent years. They are not predictions. Final cut-offs for any cycle are released by UPSC after the result and must be the source of truth.

How marks translate to rank

Civil Services ranks are decided in fine slices. Toppers in recent years have posted aggregate marks in the 1050-1100 range; the IAS cut-off often sits ~50 marks below the topper, and the broader selection list ~150 marks below. That means every five-mark improvement in answer writing can move a candidate ten to thirty places.

• Mains is where rank is built. With 1750 merit marks, a candidate who scores well in three GS papers and the Optional has a far higher rank ceiling than one who only does well in two.
• Essay and Ethics often separate ranks. These two papers reward structure, not just content. Candidates with the same factual base often score 30-40 marks apart here.
• Interview adds variance. A 275-mark interview can swing 80-120 marks between candidates. It cannot fix a bad Mains; it can finish a good one.

The CSAT 33% rule and why it matters more than people realise

CSAT being qualifying is one of the most under-prepared facts in UPSC. Three consequences follow that change strategy materially.

• A 33% miss is a wasted attempt. A candidate at 99/200 in GS but 65/200 (32.5%) in CSAT fails Prelims. The 99 counts for nothing.
• Overinvesting beyond 50% is also wasted. Marks above the 33% floor add zero. The optimal CSAT plan is "cross 50% in mocks, then stop" - the extra hours go to GS and Mains.
• A tough CSAT year is everyone's problem. When the paper turns hostile (as it has in some recent years), candidates who treated CSAT as an afterthought lose disproportionately. Two timed CSAT mocks a month is cheap insurance.