More CSIR-NET marks are lost to a poor understanding of the pattern than most candidates realise. The paper is built to reward selective, accurate attempting — and to punish guessing. Understand the structure cold, and a large slice of your score is secured before you answer a single chemistry question.
The overall structure
CSIR-NET Chemical Sciences is a single objective paper of 200 marks, taken as a three-hour (180-minute) computer-based test. It is split into three parts — A, B and C. You are shown more questions than you may attempt in each part, so part of the exam is deciding which questions to answer.
| Part | Focus | Questions / attempt | Marks each | Wrong-answer penalty |
|---|---|---|---|---|
| Part A | General aptitude | 20 given, attempt 15 | +2 | −0.5 |
| Part B | Core chemistry | ~40 given, attempt 35 | +2 | −0.5 |
| Part C | Advanced chemistry | ~60 given, attempt 25 | +4 | −1 |
That comes to roughly 75 attempts for 200 marks. Figures can be revised, so confirm against the official notification and our syllabus & pattern page.
The negative-marking maths
CSIR applies a flat 25% negative marking on wrong answers — but because the marks per question differ, the rupee value of a mistake differs too:
- Part A & B: 25% of 2 marks = −0.5 per wrong answer.
- Part C: 25% of 4 marks = −1 per wrong answer.
The expected-value logic is worth internalising. In Part C, a correct answer earns +4 and a wrong one costs −1 — a 5-mark swing per question. If you can confidently eliminate options down to a genuine two-way choice, a guess has positive expected value (0.5 × 4 − 0.5 × 1 = +1.5). A blind one-in-four guess does not (0.25 × 4 − 0.75 × 1 = +0.25, but real questions are designed with attractive wrong options, so treat that as effectively negative). The rule that follows: guess only when you've eliminated, never blind.
Sectional attempt limits
Each part caps how many questions count. Answer more than the limit and only the first batch (by the system's rule) is evaluated — extra attempts do not earn bonus marks and can cost you if they're weaker. Practically this means:
- In Part A, lock in your best 15 of 20 — leave the two trickiest aptitude items.
- In Part B, you have a small cushion (attempt 35 of ~40), so you can drop your 5 weakest.
- In Part C, the cap is the real constraint: ~60 questions but only 25 count. This is where selection wins or loses ranks.
A smart attempt strategy
- Do Part A first, fast. Aptitude is low-risk, high-speed marks. Bank 15 quick attempts and move on within ~25–30 minutes.
- Sweep Part B for certainties. Answer everything you're confident on, flag the borderline ones, and come back to fill toward your 35-attempt cushion.
- Triage Part C in passes. First pass: answer only sure-shot questions. Second pass: questions where you can eliminate to two options. Stop at 25 quality attempts — do not force extra answers just to "use" the cap.
- Protect against the penalty. A skipped Part C question costs 0; a wrong one costs 1 and the 4 you might have earned. When unsure and unable to eliminate, leave it.
- Watch the clock by part. A rough split: ~25 min Part A, ~50 min Part B, ~95 min Part C, ~10 min review.
Strategy is only reliable if you've rehearsed it. Our full-length timed mocks reproduce this exact marking and attempt structure so the decisions become automatic on exam day. For the section that decides most ranks, see our Part C strategy guide.
Rehearse the real marking scheme
Our timed mocks mirror the Part A/B/C structure and negative marking exactly. Try free sample questions first, then unlock the full bank and mocks for ₹2,999 for 6 months, then 6 months free (a full year).