Optional Subject Success Rate in UPSC: Official Data

Optional Subject Success Rate

The optional subject success rate in UPSC is the single most misread number in civil services preparation. Every year, aspirants rank optionals by “success percentage” and pick the subject sitting at the top of that list — without asking what the denominator is, how many candidates it is built on, or whether the subject caused the result at all.

This article rebuilds the picture from the primary source: Table-5 of the UPSC Annual Report, which publishes, for every optional subject, the number of candidates who appeared in the Civil Services (Main) Examination and the number finally recommended for appointment. We have taken the three most recent Annual Reports carrying this table — the 72nd (2021–22), 73rd (2022–23) and 74th (2023–24) — covering CSE 2020, CSE 2021 and CSE 2022, and pooled them into a single three-year dataset of 32,048 Mains candidates and 2,601 selections.

Quick Answer

Across CSE 2020, 2021 and 2022 combined, the overall optional-subject success rate is 8.1% — 2,601 candidates recommended out of 32,048 who appeared in Mains. Individual optionals range from roughly 4.9% (History) to 12.1% (Commerce & Accountancy) among subjects with a reliable sample size.

Six optionals beat the overall average in all three years: Commerce & Accountancy, Economics, Law, Hindi Literature, Medical Science and Chemistry. Five fell below it in all three years: History, Physics, Civil Engineering, Philosophy and Geography.

But the gap is largely a selection effect, not a subject effect. The high-rate optionals are dominated by candidates who already hold the relevant professional or postgraduate degree. Success rate tells you who takes a subject far more reliably than it tells you what to take.

What “success rate” actually means in UPSC data

UPSC does not publish a “best optional” list. What it publishes, in Section 6 and Table-5 of each Annual Report, is a two-column count per optional subject — candidates appeared in the Main Examination, and candidates recommended for appointmentwith the ratio expressed as a percentage.

The exact formula UPSC uses

The Formula

Optional success rate = (Candidates recommended for appointment ÷ Candidates who appeared in Mains with that optional) × 100

The denominator is everyone who wrote the Mains with that optional. The numerator is everyone from that pool who survived the written stage and the Personality Test and made the final list.

Two things follow immediately, and both are routinely missed.

  • It is an end-to-end rate, not an optional-paper rate. A candidate counted as “recommended” under Geography cleared Essay, four General Studies papers, both optional papers and the interview. The optional contributes 500 of 1,750 written marks — about 28.6% of the written total and 24.7% of the final 2,025 marks. The other 75% is doing most of the work.
  • It is not the same as UPSC’s other success rate. Table-4 of the same report gives a success rate by academic stream using interviewed candidates as the denominator — which is why that number sits around 40%, not 8%. Aspirants frequently quote one figure while describing the other.

The baseline you must compare every subject against

A subject’s success rate means nothing in isolation. It only means something against the all-India optional baseline for the same year.

Exam YearSourceAppeared in MainsInterviewedRecommendedMains → FinalInterview → Final
CSE 202072nd Annual Report (2021–22)10,3432,0498338.1%40.7%
CSE 202173rd Annual Report (2022–23)8,9301,8197488.4%41.1%
CSE 202274th Annual Report (2023–24)12,7752,5251,0208.0%40.4%
Three years combined32,0486,3932,6018.1%40.7%

The baseline is remarkably stable: 8.1%, 8.4% and 8.0% across the three years, averaging 8.12%. Meanwhile the interview-to-final conversion holds at roughly 40–41% throughout. In plain terms — the Mains written stage, not the interview, is where optionals decide outcomes. Roughly one in twelve Mains candidates makes the final list; two in five interviewees do.

Myth-Bust

Myth: “Subject X has a 15% success rate, so it gives me a 15% chance of clearing.”

Reality: The rate describes the historical pool that chose that subject — their degrees, their coaching access, their attempt number, their preparation years. It is a property of the candidates, not a probability transferred to you when you switch optionals.


Optional subject success rate in UPSC: complete three-year table

The table below pools CSE 2020, CSE 2021 and CSE 2022 into a single three-year figure for every optional that had at least one candidate. Pooling matters: single-year rates for smaller optionals swing wildly because they rest on a handful of candidates. The Sample Reliability column flags how much weight the number can bear — High (300+ candidates over three years), Moderate (100–299) and Low (fewer than 100, where the figure is statistically meaningless).

#Optional SubjectAppeared (3 yrs)Recommended (3 yrs)Success RateSample Reliability
1Manipuri Literature10330.0%Low
2English Literature1081715.7%Moderate
3Punjabi Literature701014.3%Low
4Management991414.1%Low
5Commerce & Accountancy6057312.1%High
6Economics6988412.0%High
7Law6107111.6%High
8Hindi Literature7258311.4%High
9Maithili Literature53611.3%Low
10Medical Science5576311.3%High
11Malayalam Literature2713011.1%Moderate
12Kannada Literature2202210.0%Moderate
13Chemistry449449.8%High
14Zoology166169.6%Moderate
15Electrical Engineering429399.1%High
16Marathi Literature2229.1%Low
17Botany9588.4%Low
18Sociology4,1683508.4%High
19Psychology479408.4%High
20Political Science & International Relations5,8124818.3%High
21Public Administration1,3911158.3%High
22Anthropology4,1703297.9%High
23Mechanical Engineering545417.5%High
24Telugu Literature8067.5%Low
25Tamil Literature12697.1%Moderate
26Statistics1417.1%Low
27Geography3,6612557.0%High
28Philosophy992697.0%High
29Civil Engineering420296.9%High
30Agriculture378266.9%High
31Mathematics1,9321296.7%High
32Gujarati Literature12486.5%Moderate
33Geology14085.7%Moderate
34Sanskrit Literature10965.5%Moderate
35Physics442245.4%High
36Animal Husbandry & Veterinary Science3825.3%Low
37History1,791874.9%High
38Urdu Literature3213.1%Low
39Bengali Literature700.0%Low
40Sindhi Literature500.0%Low
41Assamese Literature300.0%Low
42Odia Literature200.0%Low

Source: UPSC Annual Reports, 72nd (2021–22), 73rd (2022–23) and 74th (2023–24), Table-5. Rates computed by pooling raw counts across the three years; three-year totals reconcile exactly with Table-2 of each report (32,048 appeared, 2,601 recommended).

Read This Before Using The Table

Rows 1, 3 and 4 — Manipuri Literature at 30.0%, Punjabi Literature at 14.3%, Management at 14.1% — sit on 10, 70 and 99 candidates respectively across three years. Manipuri Literature’s “30%” is literally 3 selections out of 10 candidates. One extra selection would have made it 40%. These are not results; they are rounding.

Only the 20 subjects marked High carry enough candidates for the rate to be worth interpreting.

Year-by-year breakdown for every optional

The same data, disaggregated. Each cell shows recommended / appeared with the success rate in brackets, so you can see both the trend and the sample it rests on.

Optional SubjectCSE 2020CSE 2021CSE 20223-Yr Rate
Agriculture3 / 111
(2.7%)
12 / 115
(10.4%)
11 / 152
(7.2%)
6.9%
Animal Husbandry & Veterinary Science1 / 10
(10.0%)
1 / 18
(5.6%)
0 / 10
(0.0%)
5.3%
Anthropology100 / 1211
(8.3%)
90 / 1159
(7.8%)
139 / 1800
(7.7%)
7.9%
Assamese Literature0 / 2
(0.0%)
0 / 1
(0.0%)
0.0%
Bengali Literature0 / 2
(0.0%)
0 / 3
(0.0%)
0 / 2
(0.0%)
0.0%
Botany2 / 27
(7.4%)
3 / 32
(9.4%)
3 / 36
(8.3%)
8.4%
Chemistry15 / 153
(9.8%)
14 / 127
(11.0%)
15 / 169
(8.9%)
9.8%
Civil Engineering9 / 130
(6.9%)
10 / 135
(7.4%)
10 / 155
(6.5%)
6.9%
Commerce & Accountancy27 / 210
(12.9%)
21 / 140
(15.0%)
25 / 255
(9.8%)
12.1%
Economics31 / 236
(13.1%)
25 / 190
(13.2%)
28 / 272
(10.3%)
12.0%
Electrical Engineering15 / 170
(8.8%)
7 / 117
(6.0%)
17 / 142
(12.0%)
9.1%
English Literature6 / 27
(22.2%)
3 / 27
(11.1%)
8 / 54
(14.8%)
15.7%
Geography95 / 1322
(7.2%)
66 / 1079
(6.1%)
94 / 1260
(7.5%)
7.0%
Geology2 / 35
(5.7%)
1 / 29
(3.4%)
5 / 76
(6.6%)
5.7%
Gujarati Literature4 / 49
(8.2%)
2 / 32
(6.2%)
2 / 43
(4.7%)
6.5%
Hindi Literature19 / 226
(8.4%)
22 / 211
(10.4%)
42 / 288
(14.6%)
11.4%
History24 / 581
(4.1%)
25 / 574
(4.4%)
38 / 636
(6.0%)
4.9%
Kannada Literature7 / 83
(8.4%)
7 / 60
(11.7%)
8 / 77
(10.4%)
10.0%
Law23 / 167
(13.8%)
21 / 180
(11.7%)
27 / 263
(10.3%)
11.6%
Maithili Literature3 / 31
(9.7%)
2 / 12
(16.7%)
1 / 10
(10.0%)
11.3%
Malayalam Literature10 / 93
(10.8%)
8 / 74
(10.8%)
12 / 104
(11.5%)
11.1%
Management6 / 38
(15.8%)
5 / 31
(16.1%)
3 / 30
(10.0%)
14.1%
Manipuri Literature1 / 4
(25.0%)
1 / 2
(50.0%)
1 / 4
(25.0%)
30.0%
Marathi Literature1 / 9
(11.1%)
1 / 8
(12.5%)
0 / 5
(0.0%)
9.1%
Mathematics26 / 580
(4.5%)
50 / 513
(9.7%)
53 / 839
(6.3%)
6.7%
Mechanical Engineering22 / 201
(10.9%)
11 / 160
(6.9%)
8 / 184
(4.3%)
7.5%
Medical Science16 / 162
(9.9%)
24 / 196
(12.2%)
23 / 199
(11.6%)
11.3%
Odia Literature0 / 2
(0.0%)
0.0%
Philosophy20 / 347
(5.8%)
20 / 265
(7.5%)
29 / 380
(7.6%)
7.0%
Physics8 / 130
(6.2%)
7 / 131
(5.3%)
9 / 181
(5.0%)
5.4%
Political Science & International Relations154 / 1863
(8.3%)
140 / 1571
(8.9%)
187 / 2378
(7.9%)
8.3%
Psychology11 / 155
(7.1%)
15 / 127
(11.8%)
14 / 197
(7.1%)
8.4%
Public Administration38 / 509
(7.5%)
31 / 361
(8.6%)
46 / 521
(8.8%)
8.3%
Punjabi Literature3 / 25
(12.0%)
2 / 15
(13.3%)
5 / 30
(16.7%)
14.3%
Sanskrit Literature1 / 39
(2.6%)
2 / 34
(5.9%)
3 / 36
(8.3%)
5.5%
Sindhi Literature0 / 1
(0.0%)
0 / 2
(0.0%)
0 / 2
(0.0%)
0.0%
Sociology115 / 1245
(9.2%)
92 / 1087
(8.5%)
143 / 1836
(7.8%)
8.4%
Statistics0 / 5
(0.0%)
0 / 3
(0.0%)
1 / 6
(16.7%)
7.1%
Tamil Literature4 / 64
(6.2%)
2 / 27
(7.4%)
3 / 35
(8.6%)
7.1%
Telugu Literature5 / 36
(13.9%)
1 / 20
(5.0%)
0 / 24
(0.0%)
7.5%
Urdu Literature1 / 11
(9.1%)
0 / 7
(0.0%)
0 / 14
(0.0%)
3.1%
Zoology5 / 41
(12.2%)
4 / 56
(7.1%)
7 / 69
(10.1%)
9.6%

An em-dash indicates the optional had no candidate appearing in Mains that year. Of the 48 optionals UPSC offers — 25 subjects plus literature of 23 languages — only 40 to 42 recorded any candidate in a given year. Literature of Bodo, Dogri, Kashmiri, Konkani, Nepali and Santhali had no takers at all in any of the three years.


Which optionals actually perform, once sample size is respected

Filter to the 20 optionals with 300 or more Mains candidates over three years, and the noise disappears. Eleven beat the 8.12% baseline; nine fall below it.

The six optionals that beat the baseline every single year

Consistency is a far stronger signal than a single high year. These six cleared the all-India average in CSE 2020, CSE 2021 and CSE 2022:

  1. Commerce & Accountancy — 12.9%, 15.0%, 9.8% (three-year: 12.1%, 73 of 605)
  2. Economics — 13.1%, 13.2%, 10.3% (three-year: 12.0%, 84 of 698)
  3. Law — 13.8%, 11.7%, 10.3% (three-year: 11.6%, 71 of 610)
  4. Hindi Literature — 8.4%, 10.4%, 14.6% (three-year: 11.5%, 83 of 725)
  5. Medical Science — 9.9%, 12.2%, 11.6% (three-year: 11.3%, 63 of 557)
  6. Chemistry — 9.8%, 11.0%, 8.9% (three-year: 9.8%, 44 of 449)

Note the shape of this list. Four of the six — Commerce, Economics, Law and Medical Science — are subjects whose serious candidate pool is gated by a degree. A chartered accountant taking Commerce, an MBBS taking Medical Science or an LLB taking Law is not a general aspirant who picked a scoring subject; they are a specialist with three to five years of formal training in the syllabus, typically stronger academic preparation overall, and often fewer competing demands on Mains revision time.

The five optionals that stayed below the baseline every year

  1. History — 4.1%, 4.4%, 6.0% (three-year: 4.9%, 87 of 1,791). The weakest performer of any high-volume optional, at roughly 60% of the all-India rate.
  2. Physics — 6.2%, 5.3%, 5.0% (three-year: 5.4%, 24 of 442). The only major optional with a steadily falling rate across all three years.
  3. Civil Engineering — 6.9%, 7.4%, 6.5% (three-year: 6.9%, 29 of 420).
  4. Philosophy — 5.8%, 7.5%, 7.6% (three-year: 7.0%, 69 of 992). Improving, but still short of the baseline each year.
  5. Geography — 7.2%, 6.1%, 7.5% (three-year: 7.0%, 255 of 3,661). The largest optional to sit consistently below average.

The subjects UPSC itself names each year — and why they keep changing

Each Annual Report highlights the highest-scoring optionals among subjects opted by 100 or more candidates. Here is what UPSC named, year by year:

RankCSE 2020CSE 2021CSE 2022
1Law
13.8% (23/167)
Commerce & Accountancy
15.0% (21/140)
Hindi Literature
14.6% (42/288)
2Economics
13.1% (31/236)
Economics
13.2% (25/190)
Electrical Engineering
12.0% (17/142)
3Commerce & Accountancy
12.9% (27/210)
Medical Science
12.2% (24/196)
Medical Science
11.6% (23/199)
4Mechanical Engineering
10.9% (22/201)
Psychology
11.8% (15/127)
Malayalam Literature
11.5% (12/104)
5Medical Science
9.9% (16/162)
Law
11.7% (21/180)
Economics
10.3% (28/272)

Not one subject appears in all three top-three lists. Economics features in two, Medical Science in two, and everything else appears once. If the “best optional” were a stable property of the subject, the same names would recur. They do not — which is the clearest available evidence that year-to-year rankings are substantially noise, and that building a two-year preparation plan around last year’s topper subject is chasing a moving target.


The popularity trap: where everyone actually goes

Six optionals absorb two-thirds of the entire Mains pool.

Optional SubjectAppeared (3 yrs)Share of Mains PoolRecommendedSuccess Rate
Political Science & International Relations5,81218.1%4818.3%
Anthropology4,17013.0%3297.9%
Sociology4,16813.0%3508.4%
Geography3,66111.4%2557.0%
Mathematics1,9326.0%1296.7%
History1,7915.6%874.9%
Six subjects combined21,53467.2%1,6317.6%
All optionals (baseline)32,048100%2,6018.1%

67.2% of all Mains candidates take one of six subjects, and those six produce 62.7% of all selections — a slight underperformance against their share. The concentration has practical consequences that raw success rates hide.

What the concentration actually does to you

  • Evaluation is relative. With nearly 2,400 candidates writing PSIR in CSE 2022 alone, the examiner has an enormous internal distribution to calibrate against. Competent-but-ordinary answers land squarely in the middle of a very crowded band. In a subject with 150 candidates, the same relative standing is reached differently.
  • Marginal information is exhausted. Popular optionals have saturated coaching markets, meaning the baseline answer quality is high. Your competition has read the same standard sources and practised the same standard question banks. Differentiation costs more effort, not less.
  • Popularity is not a quality signal. PSIR, Anthropology and Sociology are the three most-chosen optionals, and all three sit within half a percentage point of the 8.12% baseline. Geography, the fourth most popular, sits more than a point below it. Aspirants often read popularity as crowd-verified endorsement; the data shows the crowd converges on average outcomes.

How the big six have moved across three years

  • PSIR remained the most preferred optional in all three years, growing from 1,863 to 2,378 candidates — a 27.6% rise — while its success rate drifted from 8.3% to 7.9%.
  • Anthropology overtook Geography to become the second most popular optional by CSE 2022 (1,800 candidates), with its rate easing from 8.3% to 7.7%.
  • Sociology shows the clearest single-direction decline among the big six: 9.2% → 8.5% → 7.8%, even as its candidate count grew by nearly 50%.
  • Mathematics is the most volatile high-volume optional on record: 4.5% → 9.7% → 6.3%. A candidate who chose Mathematics in 2021 on the strength of the 2020 figure, or avoided it in 2022 on the strength of 2021, was reacting to noise both times.
  • History improved from 4.1% to 6.0% but remained the lowest-performing large optional throughout, and continues to attract over 600 candidates a year.

Success rate by subject family

Grouping optionals into families removes some of the small-sample distortion and exposes a pattern individual rows obscure.

Subject FamilyAppeared (3 yrs)RecommendedSuccess Rate
Humanities & social sciences (non-literature)24,4761,9688.0%
Literature of languages (all 17)1,96720310.3%
Pure sciences3,2382307.1%
Engineering (Civil, Electrical, Mechanical)1,3941097.8%
Medical & agricultural sciences973919.4%
All optionals32,0482,6018.1%

The literature optionals as a family return 10.3% — the highest of any group, and about 27% above the all-India baseline. This is the strongest single argument for the “native language literature” strategy, and it deserves careful reading rather than enthusiasm.

Literature optionals in detail

Literature OptionalAppeared (3 yrs)RecommendedSuccess Rate
Manipuri Literature10330.0%
English Literature1081715.7%
Punjabi Literature701014.3%
Hindi Literature7258311.4%
Maithili Literature53611.3%
Malayalam Literature2713011.1%
Kannada Literature2202210.0%
Marathi Literature2229.1%
Telugu Literature8067.5%
Tamil Literature12697.1%
Gujarati Literature12486.5%
Sanskrit Literature10965.5%
Urdu Literature3213.1%
Assamese Literature300.0%
Bengali Literature700.0%
Odia Literature200.0%
Sindhi Literature500.0%
All 17 literatures combined1,96720310.3%

The mechanism behind the literature family’s performance is almost entirely self-selection by fluency. Candidates who take Malayalam, Kannada, Punjabi or Hindi Literature are overwhelmingly native or near-native speakers, frequently with formal literature backgrounds, choosing a syllabus they have absorbed since school. The barrier to entry is high enough that casual optional-shoppers never enter the pool — which is precisely why the surviving pool converts well.

Read The Small Print

Four literature optionals recorded zero selections across all three years — Assamese, Bengali, Odia and Sindhi — on a combined 17 candidates. That is not evidence they are unscoreable; it is evidence that 17 candidates over three years cannot produce a meaningful rate in either direction.

Urdu Literature illustrates the same trap from the other side: 1 selection from 32 candidates over three years (3.1%), driven by consecutive zero years in 2021 and 2022. A single selection in either year would have more than doubled the figure.

Technical optionals: the cross-domain story

Engineering optionals return 7.8% and pure sciences 7.1% — both below baseline. What makes this striking is that the recommended candidates are overwhelmingly from those backgrounds. The Annual Reports quantify the gap directly:

Among candidates recommendedCSE 2020CSE 2021CSE 2022
Held an engineering degree64.9%60.4%53.1%
Held a humanities degree23.2%25.1%29.0%
Chose an engineering optional5.5%3.7%3.4%
Chose a humanities optional (incl. literature)85.1%80.8%84.1%
Source: UPSC Annual Reports, Section 6.1 (iii), 72nd, 73rd and 74th editions.

In CSE 2022, 53.1% of successful candidates were engineers, but only 3.4% used an engineering optional. UPSC states the conclusion plainly in its own report: “most of the candidates have made a cross domain shift from their original stream to humanities.” The engineering optional success rate is therefore not measuring engineers — it is measuring the small residue of engineers who chose not to switch, a self-selected minority that is not representative of engineering aspirants at all.

One trend inside this is worth noting for its direction rather than its level: the engineering share of successful candidates has fallen steadily — 64.9% → 60.4% → 53.1% — while the humanities share has risen from 23.2% to 29.0%. The applicant profile of the examination is shifting.


Why success rate is a weak basis for choosing an optional

Problem 1: The numbers carry a margin of error that nobody quotes

A success rate estimated from a sample has a confidence interval. For small optionals the interval is so wide the point estimate is close to worthless:

Optional SubjectSample Size (3 yrs)Observed Rate95% Margin of ErrorTrue Rate Plausibly Lies Between
Political Science & International Relations5,8128.3%±0.7 pp7.6% – 9.0%
Sociology4,1688.4%±0.8 pp7.6% – 9.2%
Public Administration1,3918.3%±1.4 pp6.8% – 9.7%
Law61011.6%±2.5 pp9.1% – 14.2%
Medical Science55711.3%±2.6 pp8.7% – 13.9%
English Literature10815.7%±6.9 pp8.9% – 22.6%
Management9914.1%±6.9 pp7.3% – 21.0%
Statistics147.1%±13.5 pp0.0% – 20.6%

Management’s headline 14.1% has a 95% confidence interval running from roughly 7% to 21% — it could plausibly be a below-average optional. English Literature’s 15.7% could plausibly be 9%. Meanwhile PSIR’s 8.3% is pinned to within ±0.7 points because it rests on 5,812 candidates. The subjects with the most impressive-looking rates are precisely the ones whose rates we know least well.

Problem 2: Selection effect, not subject effect

This is the central error. Success rate measures the composition of a subject’s candidate pool, and pools differ systematically:

  • Degree-gated subjects (Law, Medical Science, Commerce & Accountancy, Economics, Management) draw candidates who have already completed three to five years of formal training in the syllabus. Their advantage is prior mastery, not the paper.
  • Language literature subjects draw near-native speakers with a decade or more of unconscious exposure.
  • Large humanities subjects (PSIR, Sociology, Anthropology, Geography) are the default choice of first-time general aspirants — including a large share of first-attempt candidates who would have a below-average conversion rate whatever they had chosen.

Switching to Law does not confer a Law graduate’s advantage. It gives you a syllabus you must build from zero, competing against people who did not.

Problem 3: The optional is a quarter of the total, and everyone clears the same GS papers

Optional Papers I and II carry 500 marks against a written total of 1,750 and a grand total of 2,025 including the interview. A candidate scoring 300/500 in the optional against a peer scoring 250/500 gains 50 marks — meaningful at the margin, but smaller than the spread routinely seen across Essay and four GS papers combined. The optional decides ranks; General Studies and Essay decide selections far more often.

Problem 4: The data is three to six years old by the time you use it

Table-5 for CSE 2022 was published in the 74th Annual Report. A candidate reading it today is projecting it onto an examination two or more cycles later, with different question papers, different examiners and a differently composed candidate pool. The volatility in Mathematics (4.5% → 9.7% → 6.3%) and Mechanical Engineering (10.9% → 6.9% → 4.3%) shows how quickly a single subject’s figure can move.

Success rate is useful as a negative filter and a sanity check, not a positive recommendation. Three defensible uses:

  • Rule out a subject you were choosing purely on hype. If History attracts 600 candidates a year and converts at 4.9%, “everyone says History is safe” deserves scrutiny.
  • Confirm a subject you already have a real edge in. If you hold an LLB, Law’s 11.6% is genuine reassurance because you belong to the pool that produced it.
  • Calibrate expectations. Knowing the all-India rate is 8.1% is a healthier anchor than any subject-level figure.

How to actually choose your optional subject

Replace “which optional has the highest success rate” with a sequence of questions that predict your outcome rather than describing someone else’s.

Step 1 — Start with genuine prior exposure

If you hold a degree in a subject on the optional list, that subject starts with an advantage no success rate can transfer to you. This is the one factor the data unambiguously supports: every consistently above-baseline optional is one where a large share of takers arrive pre-trained. Graduation subject beats every other consideration.

Step 2 — Test the syllabus, not the reputation

Download the official UPSC syllabus for your two or three shortlisted optionals and read every line. Then open the last three years of question papers for each. You are testing two things: whether you can hold interest in this material for eighteen months, and whether the question style rewards how you think — conceptual and argumentative, or factual and technical.

Step 3 — Audit General Studies overlap honestly

Overlap is real but consistently overstated in marketing. A realistic assessment:

OptionalGenuine GS OverlapWhere It Helps
PSIRSubstantialGS Paper II (polity, governance, international relations) and Essay
GeographySubstantialGS Paper I (physical geography, resources), Prelims environment and mapping
SociologyModerate to substantialGS Paper I (society) and GS Paper II (social justice, schemes), Essay
Public AdministrationModerateGS Paper II (governance) and GS Paper IV (ethics in administration)
AnthropologyModerateGS Paper I (tribal society, culture) and GS Paper II (vulnerable sections)
HistoryModerateGS Paper I (Indian and world history), Prelims art and culture
EconomicsModerateGS Paper III (economy) and Prelims economy, though at greater technical depth
PhilosophyLimited to moderateGS Paper IV (ethics) and Essay
LawLimited to moderateGS Paper II (constitution) and current legal affairs
Engineering, pure sciences, Medical ScienceMinimalMarginal support to GS Paper III science and technology
Literature optionalsMinimalEssay language quality and GS Paper I culture, for some languages

Overlap saves revision time and improves Essay quality. It does not make an unsuited subject suitable.

Step 4 — Check the infrastructure exists

Before committing, confirm that standard reference books are available and current, that previous year question papers are accessible for at least ten years, that answer copies or model answers exist to calibrate against, and that you can find at least one person who has cleared with that optional and can be asked questions. This filter alone eliminates several optionals whose success rates look attractive on paper.

Step 5 — Commit and stop comparing

The single most damaging pattern visible in aspirant behaviour is switching optionals in the second year after reading a fresh success-rate table. Switching discards a year of accumulated depth in exchange for a subject whose advantage, as shown above, is largely an artefact of who takes it. Once you have chosen on prior exposure, syllabus fit and infrastructure, the decision is closed.

Choose on fit; use the data only to veto. If your first choice sits far below the baseline and you have no prior exposure to it and it has weak GS overlap, that combination is a reason to reconsider. Any one of those factors on its own is not.


Key takeaways

  1. The all-India optional subject success rate in UPSC is 8.1% — 2,601 recommended from 32,048 Mains candidates across CSE 2020, 2021 and 2022.
  2. Among the 20 optionals with reliable sample sizes, rates run from 4.9% (History) to 12.1% (Commerce & Accountancy). Eleven beat the baseline, nine fall below it.
  3. Six optionals beat the baseline in all three years — Commerce & Accountancy, Economics, Law, Hindi Literature, Medical Science, Chemistry. Five fell short in all three — History, Physics, Civil Engineering, Philosophy, Geography.
  4. Four of the six consistent outperformers are degree-gated subjects. Their rates measure candidate quality, not paper generosity.
  5. The top-three list UPSC publishes changes completely every year. No subject appeared in all three lists across CSE 2020–2022.
  6. Six optionals absorb 67.2% of the Mains pool and return 62.7% of selections — a slight underperformance relative to their share.
  7. Literature optionals as a family return 10.3%, the highest of any group, driven almost entirely by native-speaker self-selection.
  8. 53.1% of CSE 2022 selections held engineering degrees, but only 3.4% used an engineering optional. The cross-domain shift to humanities is near-universal.
  9. Small-sample optionals carry margins of error of ±7 percentage points or more. The most impressive-looking rates are the least trustworthy.
  10. The optional is 500 of 2,025 final marks. Choose on prior exposure, syllabus fit, GS overlap and material availability — and use success rate only to veto a choice you cannot otherwise justify.

Frequently asked questions

Over CSE 2020–2022 combined, Commerce & Accountancy has the highest success rate among optionals with a reliable sample size, at 12.1% (73 recommended from 605 candidates), narrowly ahead of Economics at 12.0% and Law at 11.6%. Ignoring sample size, Manipuri Literature shows 30.0% — but that is 3 selections from 10 candidates over three years and carries no interpretive value.

The all-India average across the three most recent Annual Reports is 8.1% — 2,601 candidates recommended from 32,048 who appeared in the Main Examination. Year-wise it was 8.1% for CSE 2020, 8.4% for CSE 2021 and 8.0% for CSE 2022, making it one of the most stable figures UPSC publishes.

Among optionals with substantial candidate numbers, History has the lowest three-year success rate at 4.9% (87 recommended from 1,791 candidates), roughly 60% of the all-India baseline, and it stayed below average in all three years. Physics follows at 5.4%, and is the only major optional whose rate declined in each successive year.

PSIR is the most chosen optional in all three years, with 5,812 Mains candidates and 481 selections — a success rate of 8.3%, statistically indistinguishable from the 8.1% baseline. It is therefore an average-performing optional by conversion, and its actual case rests elsewhere: substantial overlap with GS Paper II, strong Essay support and abundant study material. Its popularity also means high baseline answer quality among your competition.

Yes, as a family. The 17 literature optionals that recorded candidates returned 10.3% over three years against the 8.1% baseline. The cause is self-selection — takers are overwhelmingly native or near-native speakers with long prior exposure, so the pool is stronger, not the paper easier. Taking a language literature you are not fluent in gains you nothing from that statistic.

In almost all cases, no. Switching costs a year of accumulated depth, and the advantage you are switching towards is largely a property of the candidates who take that subject rather than the subject itself. A switch is defensible only when you have completed less than one full cycle, genuinely cannot engage with the syllabus, or discover you hold a formal degree advantage in another listed optional that you had overlooked.

Optional Paper I and Paper II carry 250 marks each, so 500 marks in total. That is 28.6% of the 1,750-mark written examination and 24.7% of the 2,025-mark final total including the 275-mark Personality Test. The two qualifying language papers, Paper A and Paper B, are not counted in the merit calculation.

In Table-5 of its Annual Report, under the section titled “Distribution of Candidates who Appeared vis-à-vis Recommended by Optional Subjects”. Each report covers the Civil Services Examination held roughly two years earlier — the 72nd Annual Report (2021–22) carries CSE 2020, the 73rd (2022–23) carries CSE 2021 and the 74th (2023–24) carries CSE 2022. Reports are published on the official UPSC website.

Civil, Electrical and Mechanical Engineering together return 7.8%, marginally below the 8.1% baseline — but the more important number is participation. In CSE 2022, 53.1% of recommended candidates held engineering degrees while only 3.4% used an engineering optional. The overwhelming majority of engineers switch to humanities optionals, so the engineering optional rate describes a small, unrepresentative residual group.

No. The success rate is a historical property of everyone who took that subject — their degrees, preparation years, attempt numbers and coaching access. It is not a probability that transfers to you when you adopt the subject. Your own conversion depends far more on General Studies performance, answer-writing quality and preparation consistency than on which optional appears on your form.


Sources
  • Union Public Service Commission, 72nd Annual Report (2021–22), Appendix-9, Table-5 — Civil Services (Main) Examination, 2020.
  • Union Public Service Commission, 73rd Annual Report (2022–23), Table-5 — Civil Services (Main) Examination, 2021.
  • Union Public Service Commission, 74th Annual Report (2023–24), Table-5 — Civil Services (Main) Examination, 2022.
  • Supporting figures drawn from Table-2 (appeared, interviewed, recommended), Table-3 (year-wise vacancies) and Table-4 (success rate by broad stream of disciplines) of the same reports, and from Section 6.1 of each report.

All success rates in this article are computed from raw candidate counts published by UPSC. Three-year figures are obtained by pooling counts across CSE 2020, 2021 and 2022, not by averaging percentages. Pooled totals reconcile exactly with the appeared and recommended figures in Table-2 of each report.