Editorially revised on 9 October 2026.
Use a previous paper as the start of a practice and correction cycle
Previous-year question papers can help you inspect a real assessment's questions and practise responding to them. Their strategic value depends on what you do next: preserve your initial answers, verify what can be checked, identify a specific gap and revisit that gap with a clear task. Merely collecting PDFs or repeating a paper does not establish readiness or predict selection.
This guide focuses on that complete practice cycle. It does not forecast future questions, reproduce an unread paper, prescribe one preparation timeline for every exam or supply a cutoff prediction. The worked calculations below are original practice questions, not official previous-year questions.
The official UPSC previous-question-paper directory provides paper routes identified by examination and year. Its answer-key directory is a separate resource. A listed paper does not establish that a matching official key has been published. Check the actual exam, stage, paper, set and version before treating a response as officially verified.
Keep preparing
Continue with Sarkari Resume Templates₹299 — coaching के एक महीने से काफ़ी सस्ता / far cheaper than a month of coachingIdentify the paper before beginning practice
Record the examination, year, stage, paper or subject, set where relevant and official source. Read the actual paper instructions. Check that the file is complete and readable before using it. A filename supplied by a third-party site can be inaccurate or omit the version needed for comparison.
Read the applicable current syllabus and exam instructions separately. An older paper describes that paper's questions and conditions; it does not establish the current assessment pattern or eligibility. If a scheme has changed, record the difference instead of assuming the old marking or time rules still apply.
Keep the source and date of your retrieval with your practice notes. A downloaded document or archive listing is not a current application invitation. Do not infer that an official key for one examination applies to another paper simply because both appear in the same archive.
Choose a practice window and label its assumptions
A full timed attempt and a selected-question practice session are different activities. If you choose a short session, record the selected items, self-chosen time, aids used and purpose. Do not present those choices as the official exam's duration or conditions.
For example, a learner might choose a 30-minute session covering a selected set of questions to inspect their working method. That is an illustrative personal plan, not a scientifically optimal interval or mandatory exam rule. If the learner consults notes during the session, record that. It should not be compared as if it were an unaided attempt.
Preserve the first response and working before checking an answer. Otherwise a corrected answer can be mistaken for an initially correct one. A note such as “changed after reading the explanation” makes the record more useful than an unexplained final tick.
Verify answers only to the extent the evidence allows
Find an actually available matching official key when the exam provides one. Check its status, version and relevant instructions. A provisional document, revised key and final key can answer different questions. Do not treat a provisional answer as permanently settled when the authority provides a challenge or revision process.
If no matching key is available, mark the official answer as unresolved. You may work through a solution using reliable subject sources, but distinguish that review from an official key match. A coaching explanation or AI-generated answer is not an official response merely because it is confident or labelled “verified.”
For descriptive papers, a model response from a study source is not necessarily the authority's marking scheme. Review against the actual prompt, reliable subject material and available assessment guidance. Do not invent a numerical mark without a supplied rubric and qualified assessment.
Keep an error log that leads to a concrete correction
An error label is useful when it identifies what to revisit. “Careless” may be too vague. Did you misread what the question asked, use an incorrect method, confuse a unit, omit a step or leave the answer unverified? Record the evidence and a specific next task.
| Practice record field | What to preserve |
|---|---|
| Item identity | Paper, question, set or version, and source. |
| Initial response | Answer and working before checking. |
| Verification status | Matching official key, subject-source review or unresolved. |
| Specific gap | The step or interpretation that needs attention. |
| Correction task | A small worked review or original parallel exercise. |
| Later observation | What happened in the revisit, with aids and conditions noted. |
The log is a learning record, not a claim that every mistake predicts a fixed loss of marks. Apply score arithmetic only after checking the relevant scheme and response status. For the separate task of paper scoring and topic tags, see question-paper analysis.
Work through an original correction example
This original practice question says: two groups are in the ratio 3:5, and their combined total is 64. How many are in each group? The question is invented for this guide and has not been taken from an official paper.
The total number of ratio parts is 3 + 5 = 8. Each part represents 64 ÷ 8 = 8. The groups are therefore 3 × 8 = 24 and 5 × 8 = 40. Checking 24 + 40 = 64 and 24:40 = 3:5 confirms the stated relationship.
In a fictional initial response, Daya divides 64 by 3 and treats that as one group. Her working uses one ratio component instead of the total parts. The correction task is to identify the total parts before multiplying by each group's share. It does not require a judgement that Daya is bad at mathematics or a prediction about her exam result.
A second original exercise asks for the price after a 20% discount on ₹750. The discount is 0.20 × ₹750 = ₹150, so the remaining price is ₹750 − ₹150 = ₹600. In the fictional first response, Daya gives ₹150 as the final price. The error is answering the discount amount instead of the price requested.
Revisit the gap with an original parallel exercise
Daya chooses to revisit these two gaps two days later. That interval is a personal practice choice in the fictional scenario, not a claim of an optimal spaced-learning schedule. She uses new original exercises so the record can distinguish applying a method from recalling the earlier final numbers.
For a ratio of 4:7 with a total of 77, the parts total is 11, each part is 7, and the groups are 28 and 49. Their sum is 77, and 28:49 reduces to 4:7. For a 15% discount on ₹960, the discount is ₹144 and the remaining price is ₹816.
The fictional continuation states that Daya writes these working steps without notes and obtains those answers. The recorded result is two correct responses to these original revisit exercises under the stated conditions. It does not prove long-term retention, unaided performance on a real paper or likely selection. A later attempt could supply more evidence, but it should be recorded separately.
Keep full-paper practice and correction time distinct
A timed attempt records what you could do within that attempt's conditions. Correction work records what you learned or resolved afterwards. Combining the two can make a practice result look stronger than it was. Keep the original responses and correction notes separate, even when both are useful.
If time pressure affected a response, identify the actual observation. Did you leave an item unanswered, spend longer than your chosen limit or skip a check? Do not infer a psychological cause or prescribe blind guessing. Any response strategy must fit the actual marking scheme and your knowledge of the question.
Use existing mock tests when their content and format fit your preparation need. A practice score is not an official exam result. For basic worked methods, see quantitative aptitude practice.
Review the cycle rather than predicting the next paper
A useful final review asks whether the paper was correctly identified, first responses preserved, verification status honest, gaps specific and revisit outcomes recorded. It does not turn past topic frequencies into a promise that the next examination will repeat the same questions.
If a response remains unresolved, keep it unresolved and identify the source or review needed. If a corrected method works on a new original exercise, record that narrow result. If the next practice exposes another gap, revise the task. The cycle can continue without inventing a success percentage or a universal preparation deadline.
Frequently asked questions
Does every archived paper have an official key?
Do not assume that. Locate and match the actually published key, if available, and record its status.
Are the worked questions in this guide real PYQs?
No. They are original exercises used to demonstrate correction and revisit records.
Can a repeated paper score predict my selection?
No selection prediction is established here. Conditions, recall, current pattern and the actual assessment all matter.
Must I use the two-day revisit interval?
No. It is an illustrative personal choice. Choose a manageable plan and record the conditions and observations honestly.