7 ChatGPT Prompts for Exam Prep & Study Plans (2026)
Seven exam-prep prompts that turn your study sessions into active retrieval practice — diagnosing knowledge gaps, generating graded practice questions, simulating past paper-style questions, and building a focused revision schedule — instead of passive re-reading that doesn't stick.
Run = copies the prompt and opens the tool. Remix = save an editable copy to your workspace (free account).
Diagnose your knowledge gaps
🟢 ChatGPT
I have an exam in [SUBJECT] in [N DAYS]. Here's what I need to cover: [PASTE SYLLABUS TOPICS OR MODULE LIST] Ask me 10 diagnostic questions across these topics — a mix of recall, application, and explanation questions. Don't give me the answers yet. After I answer all 10, score me: identify which topics I answered confidently and correctly, which I answered incorrectly or partially, and which I couldn't answer at all. Create a 'gap map' showing: Strong / Shaky / Not covered — for each topic area.
Why this works: Diagnostic testing before studying (not after) shows you where to spend your limited time — most students re-study what they already know because it feels productive.
Good output: 10 diagnostic questions across the syllabus, then a Gap Map with three categories per topic: Strong / Shaky / Not covered.
Build a prioritized revision schedule
🟢 ChatGPT
Based on my gap map from Step 1, and given that my exam is in [N DAYS] and I have [HOURS PER DAY] to study, build a day-by-day revision schedule. Prioritization rules: - 'Not covered' topics that are likely high-exam-weight: 40% of my time - 'Shaky' topics: 40% of my time - 'Strong' topics: 20% of my time (just light review, not re-learning) - Reserve the last 2 days for full past-paper practice only - Rotate between active recall, practice problems, and written summary — no more than 90 minutes of the same activity Format as a table: Day | Time block | Topic | Activity type | Goal for session
Why this works: The 40/40/20 split is evidence-based: over-studying strong areas is the most common study-time mistake because it provides false reassurance without improving exam performance.
Good output: A day-by-day table with topic, activity type, and session goal — weighted toward gaps, not comfortable topics.
Generate practice questions for weak areas
🟢 ChatGPT
I'm weak on: [SPECIFIC TOPIC FROM GAP MAP] Generate 12 practice questions on this topic at three levels: - Level 1 (4 questions): define, list, state — pure recall - Level 2 (4 questions): apply, calculate, analyze — give me a scenario and ask me to work through it - Level 3 (4 questions): evaluate, compare, justify — exam-style questions requiring a reasoned argument Present the questions one at a time. After I answer each one, tell me: (1) whether I'm right, (2) the correct answer or reasoning, (3) the concept I need to revisit if I got it wrong. Move to the next question only when I say 'next'.
Why this works: Active retrieval (attempting to answer before seeing the answer) produces stronger memory encoding than re-reading — this is the most well-validated study technique in cognitive psychology.
Good output: 12 questions across 3 levels, delivered one at a time, with immediate feedback on each answer and a concept-to-revisit flag for wrong answers.
Simulate an exam question
🟢 ChatGPT
Simulate an [EXAM TYPE: essay / multiple choice / short answer / problem-set] exam question on [TOPIC], at the difficulty level of [YOUR EXAM: first-year undergraduate / A-level / professional certification / etc.]. Present the question as it would appear in the actual exam. Give me [TIME LIMIT] to answer — I'll tell you when I'm done. When I submit my answer, evaluate it as an examiner would: - Mark it against the likely marking criteria - Identify which marks I'd earn and which I'd lose, and why - Show me a model answer - Point out the 1–2 things that would most improve my exam technique on this question type
Why this works: Examiner-style marking (not just 'good job, you got the main ideas') shows you how marks are actually allocated — which is different from whether your answer sounds knowledgeable.
Good output: A realistic exam question, examiner-style marking of your answer with marks earned/lost, a model answer, and 1–2 technique improvements.
Create a memory aid for complex content
🟢 ChatGPT
I need to memorize: [LIST OF FACTS / PROCESSES / CONCEPTS / EQUATIONS] Create memory aids that are easy to recall under exam stress: 1. A mnemonic or acronym for any list or sequence that's 4+ items 2. A simple analogy that makes each concept feel intuitive (not just memorable) 3. A 'why does this matter' hook for each item — connecting it to something real-world makes it stickier 4. A 1-sentence summary of each item I could write from memory in 10 seconds Then quiz me: give me the first letter of each mnemonic and see if I can recall the full list.
Why this works: Analogies beat mnemonics for long-term retention because they create meaning, not just a memory hook — you can reconstruct them from first principles even if you forget the mnemonic.
Good output: Mnemonics, analogies, real-world hooks, and 10-second summaries for each item — plus a retrieval quiz at the end.
Teach the concept back
🟢 ChatGPT
I'm going to teach you [CONCEPT] as if you're a student who has never heard of it. Interrupt me, ask follow-up questions, and push back if I say something vague or incorrect — act like a curious but skeptical student. After I finish explaining, give me feedback: 1. What did I explain accurately and clearly? 2. What did I miss or get wrong? 3. Where was I vague — a sign I don't fully understand it yet? 4. What would a top student add that I didn't mention? I'll start: [START YOUR EXPLANATION]
Why this works: The Feynman technique — teaching a concept to expose gaps — is one of the strongest known study methods. Having ChatGPT actively push back (not just listen) makes the gaps visible in real time.
Good output: An interactive teaching session where the model plays skeptical student, followed by structured feedback on accuracy, completeness, and conceptual gaps.
Last-day rapid review
🟢 ChatGPT
It's the day before my exam. I don't have time to re-study anything deeply. For [SUBJECT / TOPICS LIST], give me: 1. The 10 most likely exam topics or question types (based on common exam patterns for this subject) 2. A 1-sentence definition or rule for each 3. The 3 most common mistakes students make in this exam, and how to avoid them 4. A 5-minute active recall drill: ask me 5 rapid-fire questions, I answer from memory, you tell me immediately if I'm right 5. A confidence-calibration check: ask me to rate my confidence (1–5) on each topic — flag any I rated high but might be overconfident on
Why this works: Last-day review should be recall activation, not learning — this prompt uses everything you already know rather than trying to add new knowledge, which consolidates better overnight.
Good output: 10 likely exam topics + 1-sentence rules, 3 common-mistake warnings, a 5-question rapid recall drill, and a confidence calibration check.
Frequently Asked Questions
How early before an exam should I start this sequence?
Step 1 (diagnostic) works best 2–3 weeks before the exam, giving you time to address real gaps. Steps 2–6 can run continuously through your revision period. Step 7 is designed for the final 24 hours.
What subjects do these prompts work best for?
Steps 1, 2, 3, and 4 work for any subject where the exam has structured questions. Step 5 works best for content-heavy subjects (history, biology, law, business). Step 6 works best for concepts that can be explained, not just recalled. Step 7 is universal.
Is ChatGPT's content always accurate for my specific course?
Not always — the questions and model answers it generates reflect common academic knowledge, which may not exactly match your course's syllabus or your lecturer's specific terminology. Always cross-reference with your course notes. Use ChatGPT for the practice format and feedback process, not as your primary source of content.
Make this workflow yours
Remix this pack into your workspace — edit the prompts, track your runs, and never hunt for them in old chats again.