NEET negative marking: when to guess, with the maths worked out

In NEET, guessing pays off on average whenever your chance of being right is above 20%. The marking is +4 for a correct answer, −1 for a wrong answer and 0 for an unanswered question (NTA NEET (UG) 2026 Information Bulletin, https://cdnbbsr.s3waas.gov.in/s37bc1ec1d9c3426357e69acd5bf320061/uploads/2026/02/202602231394640855.pdf), so the expected value of a guess is 5p − 1 marks, where p is the probability you pick correctly. That is positive for any p above 0.2.

The complication is that your real chance on a hunch is not the same as a random 25%. Below is the maths, the reason real guesses underperform it, and a rule you can actually use in the exam hall.

The expected-value maths

If you answer a question and your chance of being correct is p, you gain 4 marks with probability p and lose 1 mark with probability 1 − p. Your expected score is 4p − 1 × (1 − p) = 5p − 1. Leaving it blank scores 0. So a guess is worth making, on average, when 5p − 1 is greater than 0, which means p greater than 0.2.

• Random pick from 4 options (p = 0.25): expected value +0.25 marks.
• One option confidently eliminated (p = 0.33): expected value about +0.67 marks.
• Two options eliminated (p = 0.5): expected value +1.5 marks.
• Hunch that is right one time in five (p = 0.2): expected value 0.
• Hunch that is right one time in ten (p = 0.1): expected value −0.5 marks.

So is blind guessing actually good?

On paper, a genuinely random guess among four options has a small positive expected value, +0.25 marks. You will often hear that blind guessing in NEET has negative expected value. With +4/−1 marking and four options, that is not true of a truly random pick.

But 'small positive on average' comes with large swings. Twenty random guesses have an expected total of +5 marks, and roughly a 23% chance of a net loss, because you need at least four right out of twenty to break even. Two-option guesses behave far better: ten of them have an expected total of +15 marks and only about a 1% chance of a net loss.

More importantly, your guesses in the exam are not random, and that changes everything.

Why your real guesses are worse than random

NEET options are written so that the wrong ones are attractive. A good distractor is the answer you get from a common misconception, a missing factor, a swapped term or a half-remembered NCERT line. When you guess from a vague feeling, you are more likely to be pulled towards exactly those options than a coin would be. That can push your real hit rate on 'gut' guesses below 20%, where guessing loses marks.

Pressure makes it worse. Students tend to guess more when they are anxious or short of time, and on the question types where distractors are strongest, such as assertion-reason and multi-statement questions.

Measure your own guessing accuracy

You do not need to guess about guessing. In every mock, mark each answer you are not sure of as either a 'two-way' guess (you eliminated two options) or a 'hunch' (you could not eliminate two). After ten mocks, compute your accuracy in each category.

If your hunch accuracy is above about 25%, hunches are roughly paying their way. If it is below 20%, they are costing you, and you should leave those questions blank. Most students find their two-way guesses are comfortably profitable. The numbers are yours, which makes them far more useful than any general rule, including this article's.

A practical guessing rule

Until you have your own data, this rule is a reasonable default. It is built on the maths above and on the fact that eliminations are more reliable than hunches.

• Eliminated two or more options with genuine reasons: answer.
• Eliminated one option with a genuine reason: answer if you have any positive lean; otherwise consider leaving it.
• Eliminated nothing, pure hunch: leave it blank unless your own mock data shows your hunches beat 20%.
• Assertion-reason and multi-statement questions you cannot resolve: leave blank more readily, because their distractors are the most convincing.
• Never guess just to 'use up' remaining time. Use it to recheck questions you have answered.

The tie-breaker cost of wrong answers

Wrong answers also matter beyond the −1. NTA's tie-breaking rules for NEET (UG) 2026 resolve equal scores first by Biology marks, then Chemistry, then Physics, and then by the candidate with a lower proportion of incorrect to correct answers across all subjects. That is a small cost in a rare situation, but at NEET's density many candidates share a score, so it is worth knowing that a careless guess can count against you twice.

The bigger picture: guessing is a small lever

Guessing policy is worth a few marks either way. Knowing more, forgetting less and making fewer careless errors are worth far more. The best way to reduce negative marking is to convert hunches into eliminations through practice, so that more of your uncertain questions become two-way choices.

Practising with questions that explain every wrong option helps directly: you learn which distractors catch you and why. The free NEET question bank at lumineet.com/questions is built that way, with chapter-wise original MCQs and an explanation for each distractor.

A worked example across a whole paper

Here is how the rule plays out over a full paper. Suppose you finish with 150 questions you are confident in, 20 where you have eliminated two options, and 10 pure hunches. Say you get 140 of the 150 right. That is 560 − 10 = 550 marks from those.

The 20 two-way guesses have an expected value of 20 × 1.5 = +30 marks. If you answer them, you most likely end up somewhere around 580. The 10 hunches are the uncertain part. If your personal hunch accuracy is 25%, answering them adds about 10 × 0.25 = +2.5 marks on average. If it is 10%, answering them costs about 10 × 0.5 = 5 marks on average. The difference between those two cases is only a few marks, which is why the two-way guesses matter more than the hunches.

The practical lesson is to spend your review time converting hunches into two-way guesses. Eliminating even one more option on a doubtful question does more for your score than any rule about how many questions to leave.

Related on Lumi

• The 3 silent mistakes costing NEET aspirants 40+ marks
• Why you score 600 in mocks and 500 in the real NEET
• How to analyse a NEET mock test in 30 minutes: a protocol and an error log
• Silly mistakes in NEET Physics: a taxonomy, and a fix for each