The latest published JEE Main framework uses two sessions and the applicable best NTA score under its result rules. That creates two preparation checkpoints, but NTA has not yet published the 2027 bulletin. Use this roadmap as a planning baseline and update the dates, pattern and result rule when the new bulletin appears.
The exam, precisely
- 75 questions, 300 marks, 3 hours, computer-based (Paper 1, B.E./B.Tech).
- 25 questions per subject (Physics, Chemistry, Mathematics): 20 MCQs + 5 numericals with no internal choice. Unanswered questions receive 0.
- +4 correct, -1 wrong in both sections, including numerical (Section B) questions. (Full pattern and marking maths.)
- Two sessions in the latest published cycle; the applicable best NTA score is used under that bulletin's result rules. Confirm the 2027 schedule and rule on release.
A two-session planning framework
Treat the sessions as a system, not two isolated exams:
- Session 1 = a complete attempt and diagnostic. Prepare seriously. Its subject/shift performance, error log and exam-day execution become evidence for the next block.
- Between sessions = targeted repair. Rank weak chapters and behaviours by marks lost, recurrence and time needed. Repair a small number at a time and retest them.
- Session 2 = another full attempt. The goal is not automatically "peak later"; it is to arrive with specific Session 1 defects corrected while retaining strong areas.
Do not treat Session 1 as a throwaway practice attempt. A strong first result can reduce pressure, while a weaker result can still produce a precise repair plan. Prepare seriously for both sessions if the 2027 bulletin retains that structure.
Subject plan
Physics: combine concept selection, units and numerical execution; classify errors before deciding whether the fix is theory or practice.
Chemistry: maintain separate workflows for Inorganic recall, Organic mechanism/reaction reasoning and Physical Chemistry numericals. Use the official syllabus and current materials rather than a universal "easy subject" assumption.
Mathematics: build method recognition and written execution through topic-wise problems and mixed timed sets. Let official papers and your accuracy decide chapter order instead of copying a static weight list.
Across all three, a useful default is one core resource per subject, previous-year questions after each topic, and an error notebook that informs later revision.
A 12-month starting shape
- Months 1-6 — foundation: cover the syllabus in parallel across all three subjects, with PYQs after each topic. Move on only after a first-pass explanation, representative problems and an error review — not after merely watching a lesson.
- Months 7-9 — integration: close remaining first-pass gaps, introduce mixed tests and increase full-syllabus exposure. The transition signal is that every official syllabus unit has been studied once and your test log can name the remaining weak areas.
- Final 4-6 weeks before Session 1 — consolidation: taper new material, stabilise subject order and timing, and rehearse the published format. Use calendar weeks left, not a copied mock count, to set frequency.
- Official gap between sessions — repair: rank Session 1 losses by marks, recurrence and repair cost; correct a bounded set, retest it, then run full papers to ensure strong areas have not decayed.
- Final weeks before Session 2 — retention: reduce resource switching, revisit the error log and maintain full-paper execution.
Treat these durations as a starting allocation, not a law. Compress or extend a block using its transition evidence, and map all dates onto the official 2027 schedule when published.
A repeatable study week
Use subject-parallel concept/problem blocks, one mixed test, one full or part-syllabus test appropriate to your stage, and a review block that updates the error log. Every week should produce evidence: topics closed, PYQs attempted, errors corrected and a named next priority.
Avoid carrying a vague "backlog". Convert it into chapter-topic tasks with a reason and deadline. If a chapter repeatedly rolls over, reduce its scope and finish one testable unit.
Mock discipline
Increase full-paper frequency only when there is enough time to review and correct each attempt. Track marks, accuracy, attempts, subject balance, time loss and repeated error categories. A private mock percentile is provider-specific and must not be presented as an official NTA percentile.
After Session 1, separate syllabus gaps from execution gaps. A low Mathematics score caused by six unfinished solvable questions needs a different repair from one caused by missing Calculus concepts.
Quick answers
Should Session 1 be treated as practice? No. It is a real attempt. Its diagnostic value is additional, not a reason to prepare casually.
Will Session 2 always be higher? No. Improvement depends on correcting specific weaknesses while retaining previous strengths.
How often should I take full mocks? Use a cadence that leaves time for complete review and targeted correction. Increase frequency near the exam only if review quality remains intact.
Should I track mock percentile? Treat it as a provider/cohort metric. Track your own marks, accuracy and errors consistently; official NTA percentile is calculated from examination shift data.
