Negative marking is the most misunderstood aspect of CLAT. Students either over-attempt and lose marks to penalties, or under-attempt and leave easy marks on the table. This guide gives you a probability-based framework to make the right call on every question.
Before you can strategise around negative marking, you need to know the exact rules. CLAT follows a straightforward marking scheme that has remained consistent for several years. Here are the numbers you need to memorise:
The maximum possible score is 150 marks (all 150 questions correct). There is no fractional marking for partially correct answers — each question is binary: correct (+1) or incorrect (-0.25). Unanswered questions score zero and carry no penalty whatsoever.
Net Score = (Correct × 1) − (Incorrect × 0.25)
Example: A student attempts 120 questions, gets 95 correct and 25 wrong. Net score = 95 − (25 × 0.25) = 95 − 6.25 = 88.75 marks. The 30 unanswered questions contribute nothing — neither positive nor negative.
This -0.25 penalty is mild by exam standards. Many competitive exams use -0.33 or even -0.50 per wrong answer. The relatively low penalty in CLAT means the mathematical threshold for attempting a question is lower than most students think. Understanding this is the foundation of every strategy that follows.
Most students rely on gut feeling to decide whether to attempt a question. This is wrong. There is a precise mathematical framework — expected value — that tells you exactly when attempting is profitable. Here is how it works:
Expected Value (EV) is the average outcome you would get if you faced the same decision many times. If the EV of attempting a question is positive, you should attempt it. If it is negative or zero, skip. Let us calculate the EV for different levels of elimination:
Slightly positive — but high variance. Not recommended as a strategy.
Clearly positive. Attempt the question.
Strongly positive. Definitely attempt.
If you can eliminate even one option, attempt the question. The expected value is clearly in your favour at +0.167 marks per question. Over 20 such questions in an exam, that is an extra 3.3 marks — enough to shift your rank by several hundred positions.
A surprising fact: even a pure random guess — with zero elimination — has a slightly positive expected value of +0.0625 at the -0.25 marking scheme. Mathematically, guessing every unanswered question would give you a small expected gain. However, this does not mean you should guess everything blindly. The reason is variance.
Over 150 questions, random guessing adds significant noise. In any single exam sitting, you could get lucky (gaining 5-8 marks) or unlucky (losing 5-8 marks). Since you only sit for CLAT once, the variance risk is real. Expected value is a guide for when you have some basis for elimination, not a licence to guess randomly. Use the framework where you can eliminate at least one option — that is where the math becomes both profitable and reliable.
How does your attempt rate and accuracy rate interact to produce your final net score? Let us model three common student profiles and see which approach actually yields the highest score:
The balanced approach (135 attempted at 80% accuracy) yields the highest net score at 101.25. Going too conservative leaves marks on the table. Going too aggressive drops accuracy faster than it adds marks. The sweet spot is 130-140 attempts with 78-82% accuracy. Find your personal sweet spot through mock data.
Notice the penalty column. Even in the aggressive scenario, the total penalty is only 9.06 marks — roughly 6% of total marks. Negative marking is not the score-killer students imagine. The real cost is the 15-30 marks students leave on the table by not attempting questions they could have gotten right. Under-attempting is almost always more costly than over-attempting with reasonable confidence.
Not all sections carry the same negative marking risk. Your attempt strategy should differ based on the nature of each section. Here is a section-by-section breakdown:
Passage-based, so accuracy tends to be high. The answer is almost always derivable from the passage text. Attempt all questions unless you are completely lost on a passage. If unsure about one question within a passage, skip that individual question — do not skip the entire passage.
This is the most dangerous section for negative marking. Although passage-based, GK tests factual knowledge — you either know the fact or you do not. Elimination strategies work poorly here because all four options often look plausible if you do not know the underlying fact. Skip questions where you genuinely have no idea. Only attempt if you can identify the fact being tested.
High attempt rate recommended. The passage provides a legal principle and the question asks you to apply it to a fact pattern. The answer is almost always derivable from careful reading. Skip only if the principle itself is genuinely confusing or the fact pattern is extremely ambiguous. This section rewards careful reading, not prior knowledge.
Similar to Legal Reasoning — passage-based with answers derivable from the text. Most questions test inference and argument analysis. Attempt most questions. Skip only time-consuming arrangement or sequencing questions that would eat into your time budget for higher-value sections.
If you know the concept, attempt. If you can estimate the answer range, attempt and eliminate. Do not waste time solving to exact values when estimation can narrow it to two options. Skip only if you have zero idea how to approach the problem. QT has the fewest questions, so each mark matters more — but so does each time minute spent here.
The pattern is clear: passage-based reasoning sections (English, Legal, Logical) carry low negative marking risk because the answer is in the passage. Knowledge-based sections (GK) carry the highest risk. Calculation-based sections (QT) sit in the middle. Adjust your skip threshold accordingly — be aggressive in reasoning sections, selective in GK.
After every mock test, you should not just check your score — you should categorise every wrong answer into one of four types. Each type has a completely different fix, and treating them all the same is why most students never reduce their negative marking loss.
You knew the correct answer but marked the wrong option. Misread the question, selected the wrong bubble, or confused "not" with "except". Fix: slow down on easy questions. Read the question stem twice. These are the most frustrating errors because they are entirely preventable.
You eliminated the correct option and chose from the remaining wrong options. This happens when your passage reading is superficial or you misunderstand a legal principle. Fix: go back to the passage and identify exactly where your reasoning went wrong.
You had no idea and guessed randomly. The guess happened to be wrong. Fix: tighten your skip criteria for this question type. If you are consistently guessing wrong in a particular area, that area needs content revision, not more guessing.
You attempted a question on a topic you had never studied or a current affairs fact you had never encountered. Fix: this is a content gap, not a strategy problem. Add the topic to your revision list. For GK, add it to your daily current affairs reading.
Track your negative marking loss per mock test. Create a simple spreadsheet: date, total penalty marks lost, breakdown by category (A/B/C/D), and breakdown by section. Over 10-15 mocks, you will see clear patterns. Maybe 40% of your negatives come from careless mistakes in Legal Reasoning — that is a highly specific, fixable problem.
Your total negative marking loss should decrease over time. If it is not decreasing, you are practising without analysing. The error log is the single most effective tool for reducing negative marking — more effective than any "tips and tricks" article. Use our mock analysis guide for a complete framework.
After analysing thousands of mock test performances, these are the most common negative marking mistakes — and each one is avoidable:
Students who attempt all 150 questions without a skip framework typically lose 15-20+ marks to negatives. Their net score is often lower than a student who attempted only 130 questions with higher accuracy. Blind aggression is not a strategy.
The opposite extreme. Students so afraid of -0.25 that they leave 30-40 questions unanswered. Even at 60% accuracy on those questions, attempting them would have added 10-15 net marks. Fear of negative marking costs more than negative marking itself.
Taking mock after mock without analysing where your penalties come from. If you do not know whether your negatives are from GK guessing or Legal Reasoning carelessness, you cannot fix the problem. Data-less practice is wasted practice.
The classic time management trap. You spend 4 minutes on a single tough question, then panic-guess on the last 10 questions. Those 10 panic-guesses at 25% accuracy cost you roughly 1.875 marks in penalties. The single question was not worth 4 minutes.
The theme across all four mistakes is the same: they stem from emotion, not analysis. Fear, overconfidence, and poor time management are emotional problems disguised as strategic problems. The fix is always the same — collect data from your mocks, analyse it, and let the numbers guide your attempt strategy. Negative marking is a math problem, not a feelings problem.
Yes. The -0.25 penalty per incorrect answer applies uniformly across all five sections — English, Current Affairs/GK, Legal Reasoning, Logical Reasoning, and Quantitative Techniques. There is no section where incorrect answers are penalised more or less heavily. The penalty structure is the same for every single question on the paper.
AILET (NLU Delhi) does not have negative marking as of 2026 — every unanswered or incorrect question simply scores zero. LSAT India also has no negative marking. SLAT (Symbiosis) uses -1 for every 4 incorrect answers. MH CET Law has no negative marking. Always check the latest exam notification since marking schemes can change year to year.
There is no universally "safe" number — it depends on your accuracy rate. However, data from top rankers shows most attempt 130-140 of 150 questions. If your mock accuracy is consistently above 75%, aim to attempt 135+. If your accuracy is closer to 65-70%, keep your attempts around 120-125 until accuracy improves. The key metric is net score, not attempt count.
Yes. The expected value math strongly favours attempting when you can eliminate even one option. With one option eliminated, you have a 1 in 3 chance of being correct. The expected value is (1/3)(+1) + (2/3)(-0.25) = +0.167 marks. Over 10 such questions, you gain roughly 1.67 marks compared to leaving them blank. Always attempt when you can eliminate at least one option.
For the average CLAT aspirant attempting 120-130 questions, negative marking costs between 4-8 marks. Students who attempt aggressively without strategy can lose 12-15 marks. Top rankers typically lose only 3-5 marks to negatives because their accuracy is high and they use elimination effectively. Track your negative marking loss in every mock — it should trend downward over time.
Technically, yes. With 4 options and -0.25 marking, the expected value of a pure random guess is +0.0625 marks per question. However, this is a statistical average over a very large number of questions. On a single exam with 150 questions, the variance from random guessing can easily swing your score by 5-10 marks in either direction. Random guessing is not a reliable strategy — informed elimination is.
Use the error log method. After every mock, categorise each wrong answer: careless mistake, incorrect elimination, pure guess, or genuine lack of knowledge. Each category has a different fix. Careless mistakes need slower reading. Incorrect elimination needs better passage analysis. Pure guesses need stricter skip criteria. Track your category-wise breakdown across mocks and target the biggest loss bucket.
Yes. In the first half of the exam (first 60 minutes), be slightly more selective — you have time to think carefully and should use it. In the last 15-20 minutes, if you have unanswered questions and can eliminate even one option, attempt them. The opportunity cost of leaving questions blank is highest at the end when you cannot come back. Many students leave 10-15 marks on the table in the last 10 minutes by being too conservative.