How to Use Mock Tests for JEE Main Preparation: Selection, Numerical Questions and Two Sessions

The short answer: in JEE Main, use mock tests to practise selection. The paper has 75 compulsory questions in 180 minutes (as reported for 2026), they differ widely in how long they take, and a wrong answer costs a mark, including in the numerical-value section. The candidates who gain most from mocks are those who record, for each question, whether they attempted it, how long it took and whether it was worth it, and then adjust a written selection rule before the next paper.

Why selection matters more than coverage

At 75 questions in 180 minutes the average is 2 minutes 24 seconds per question, but the spread is wide: a recall question in Chemistry takes under a minute, while a Mathematics problem with a long calculation can take five. If you try to solve the paper in order, you will spend your best time on whatever comes first, and run out of time before reaching easier questions later. A mock is the safe place to learn which questions to take, which to leave and which to return to. The structure is described in the JEE Main exam pattern guide.

Build a selection ladder

A selection ladder is a short written rule that says what you do with a question depending on what you see in the first fifteen seconds:

Write your own version on a card. After each mock, check how often you followed it and what happened when you did not. Candidates are often surprised by how many marks they lost to breaking their own rules, usually on a question they were "close to solving".

What to record in every mock

Keep a simple log with one row per subject:

• Attempts, correct, wrong MCQ and wrong numerical answers.
• Time used, and time left at the end.
• The one question that took the most time, and whether it paid off.
• Net score and marks per attempted question.
• The change you tested (a new sequence, a different cap).

A worked example: three papers, one decision

Suppose a candidate records the following for Mathematics across three papers (illustrative figures):

In paper 1 the candidate attempted 22 of 25 questions at 59% accuracy and took 82 minutes, which squeezed the time for the other subjects. In paper 2 they attempted four fewer questions, scored four more marks and saved ten minutes. In paper 3, with one more attempt and the same time cap, they scored nine more marks. The pattern is a candidate learning to leave the right questions. The decision this supports: keep the same time cap and test a small relaxation, such as allowing one extra question in the second pass. If accuracy stays above about 75%, keep it; if it falls, go back. The ten minutes saved per paper in Mathematics are time that can go to Physics, where the same candidate might be running short.

Handling the numerical-value section

The five numerical-value questions per subject are 60 of the 300 marks. They are compulsory in the current pattern, and reports on the 2026 bulletin say they carry a −1 penalty. That changes how you should take mocks:

1. Use a mock that applies the same penalty. A test that does not penalise numerical answers will encourage guessing that would cost marks in the real exam.
2. Do the numerical questions after the MCQs of the subject, not at the start; by then you know how much time is left.
3. Sanity-check before typing: size, sign, units and any rounding instruction.
4. Keep a numerical log: attempted, correct, and the cause of each error (method, arithmetic or format).

The expected value of a blind numerical guess is close to −1, since the chance of a lucky match is tiny. So the rule is simple: attempt only with a computed answer. For a few examples of what these questions look like, see the Mathematics, Physics and Chemistry pages.

Use the time-per-question table to find your pace

For each subject, estimate your average time per question type from your mocks. Then compare it with a target plan. If, for example, your Chemistry MCQs take 90 seconds on average and the plan says 45, your Chemistry is where time is being lost. Candidates often assume Mathematics is the problem when Chemistry recall questions are actually the slowest relative to their value. The JEE Main mock test strategy includes a reference table of typical times.

Sequence experiments

Test your subject sequence like any other variable. For example, run Chemistry, Physics, Mathematics for two papers, then Mathematics, Chemistry, Physics for two papers, and compare net score and time left. If one sequence gives a clear gain, adopt it and rehearse it. Do not change it in the last week.

Planning around the two sessions

JEE Main is held in more than one session, and the better score can generally be used. That suggests a plan: treat the first session as a full attempt, with your best preparation; use the weeks between sessions to repair the chapters its result exposed; and retake weekly full mocks in the last fortnight before the second session. Do not use the second session as a "free" attempt without preparation, because your attention and nerves will be tested twice. The details of how scores combine are in the NTA information bulletin.

Mistakes to avoid

• Taking a mock with no penalty on numerical questions and learning to guess.
• Solving questions in order and spending the best time on the first few.
• Reviewing only wrong answers. Questions you solved but slowly, and ones you guessed correctly, are future errors.
• Looking at mock ranks as predictions. A rank compares you with that test's takers, not with the national pool.
• Adding more mocks when scores stall. Add a repair session instead.

A checklist for each mock

1. Write your selection ladder and time plan before you start.
2. Take the paper under real conditions, in your session's time slot.
3. Record attempts, time and net score per subject.
4. Classify every wrong answer: concept, calculation, reading, selection, guess or pressure.
5. Redo skipped questions untimed.
6. Update your error log and write one change for the next paper.

To begin, try a free JEE Main mock test or browse the JEE Main hub. Please confirm the pattern and session details in the official NTA bulletin; tests on TestMint are independent practice papers.