The GMAT review and edit feature is one of the most misunderstood mechanics on the Focus Edition — it looks like a safety net, but used carelessly it can quietly cost you points. You can bookmark as many questions as you like in each section, yet you can only change three answers, and the screen vanishes the moment time runs out. This guide breaks down exactly how the feature works, when to flag, when to edit, and the common mistakes that turn a helpful tool into a score liability.
The GMAT review and edit feature is built into every section of the GMAT Focus Edition. Mechanically, it is one specific screen that only appears under very specific conditions — and a bookmark tool you use while you work. Getting these two pieces straight removes about 80% of the confusion students have on test day.
The Question Review & Edit screen is a numbered list of every question in the section you just finished. From that list you can jump back to any question — not only the ones you flagged — see your current answer, and change it if you want to. The screen only appears after you submit an answer to the final question in the section, so you cannot access it mid-section.
As you work through each section, the GMAT Focus bookmark flag lets you mark any question for easy navigation later. Bookmarks are unlimited: you can flag one question or all of them, and flagged items show a bookmark icon on the review screen so they are easy to find. Crucially, the bookmark is a navigation aid only — it does not reserve an edit, and you are never forced to change an answer on a bookmarked question.
If the section clock hits zero before you answer the final question — or the instant after — the Review & Edit screen is skipped entirely. There is no partial review, no grace period, and no way to undo it. Any bookmarks you placed for that section become useless because you never get a chance to revisit the questions. That single rule is why pacing matters more than flagging for most students.
This is where most students get the rules wrong — and where GMAT change answers questions come up most often in forums. The cap sounds simple, but a single word matters: changes, not questions. That distinction decides whether your three edits help you or vanish on a single indecisive answer.
You are allowed a maximum of three answer changes per section. Every time you switch an answer — during the section or on the Review & Edit screen — it counts as one edit against that section's budget. If you change the same question's answer twice, you've used two edits on one question, and your final answer might still not be correct.
The GMAT Focus Edition is a question-adaptive test. The scoring algorithm assigns later question difficulty based on how you answer the earlier ones, and it needs most of your responses to stay fixed to produce an accurate score. Unlimited edits would let test-takers effectively rewrite their performance after the fact, breaking the model. Three edits is the compromise: enough to catch a careless error, not enough to override your real ability.
The cap resets each section. That means up to 3 edits in Quantitative Reasoning, 3 in Verbal Reasoning, and 3 in Data Insights — for a theoretical maximum of 9 GMAT 3 edits per section across the whole exam. Unused edits do not roll over to the next section. If you only need one edit in Quant, the other two disappear when the section ends.
Worked Example
Setup: Halfway through the Verbal section you change your answer on Question 9 from B to D. Ten minutes later, on the review screen, you decide D was wrong and switch it back to B. How many of your three edits have you used?
A good GMAT edit answer strategy starts before you ever reach the review screen. Every flag you place costs you time to navigate to and away from later. The question isn't "can I flag this?" — bookmarks are unlimited — but "is this flag realistically useful if I only have 2 minutes left?"
The cleanest flagging trigger is the two-candidate scenario: you eliminated three answer choices with real reasoning, narrowed it to two, and had to guess between them under time pressure. That kind of question is already 70% solved — a 20-to-30-second revisit often tips you to the correct one because your work is already on the page. Flag it, guess your best option, and move on.
Do not flag questions you never cracked — the ones where you'd need to re-read the entire prompt from scratch. Re-reading eats your section clock and rarely changes the result. Also do not flag questions you are confident about. Second-guessing confident answers at the review screen is one of the most common ways test-takers drop points they already earned.
A realistic flagging budget is 3–6 questions per section for most students. With roughly 2 minutes of review time at the end, flagging 9–10 questions means fewer than 15 seconds each — which is not enough to reason carefully. Set a cap before the section starts and respect it.
Worked Example
Setup: On a Data Insights Graphics Interpretation problem, you narrow the answer to two choices in under a minute but can't commit. The rest of the problem would take another 90 seconds to re-read.
| Scenario | Flag It? | Edit It? | Reason |
|---|---|---|---|
| Narrowed to 2 choices, had to guess | Yes | Maybe — only if you can confirm a clear reason to switch | Quick revisit can tip you to the correct answer |
| You solved it, you're 90%+ confident | No | No | Second-guessing confident answers is where scores drop |
| You'd need to re-read the entire prompt | No | No | Re-reading eats the clock and rarely changes your answer |
| You picked A, then saw something that points to C on the next question | Yes | Yes — if the reasoning is concrete, not a hunch | This is exactly what the 3 edits are for |
| You're already behind pace | No | No | Protect the section clock instead of planning reviews that may not happen |
| You remembered a formula or rule you forgot | Yes | Yes | New information is the cleanest edit trigger |
Flagging and editing decisions happen in seconds. These two tools turn the rules above into a quick lookup you can practice with on mocks so the right call becomes automatic on test day.
Pick the situation you're in mid-section and see whether to flag, edit, or move on.
Enter how many questions you bookmarked and roughly how long a realistic revisit takes you. The tool returns the minimum section-clock buffer you need, whether your flag count fits that buffer, and how many edits to cap yourself at.
A surprising number of students treat the Review & Edit feature as a free do-over. It isn't — and understanding why is the difference between using the feature for modest gains and actively burning points with it.
The GMAT Focus Edition is question-adaptive. As you move through a section, the algorithm uses your prior responses to select the next question's difficulty. If you get the first several questions wrong, you push yourself into an easier pool for the rest of the section — and an easier pool caps the ceiling on your section score.
Changing a wrong early answer to the correct one at the review screen does not re-run the section. The questions you already saw were selected based on your initial performance, so the difficulty path you walked is locked in. You can keep the edit's credit on that one question, but the harder questions you never reached are gone.
The practical consequence is simple: play the first half of each section as if the feature didn't exist. No sandbagging, no mental IOUs, no "I'll fix it later." Reserve the idea of editing for the back half of the section and for the review screen, where corrections can help without silently dropping your difficulty ceiling.
All three GMAT Focus sections share the same 3-edit cap, but their timing realities are different. Question counts, question types, and per-question averages should shape how many flags you realistically use in each one.
| Section | Questions | Time | Avg. Time / Question | Edit Limit | Suggested Max Flags |
|---|---|---|---|---|---|
| Quantitative Reasoning | 21 | 45 min | ~2:08 | 3 | 4–6 |
| Verbal Reasoning | 23 | 45 min | ~1:57 | 3 | 3–5 |
| Data Insights | 20 | 45 min | ~2:15 | 3 | 3–4 |
| Exam total | 64 | 2 hr 15 min (plus one 10-min break) | — | 9 total | 10–15 across the exam |
With roughly 2 minutes and 8 seconds per question, Quant gives you the most breathing room of the three sections. Problem-solving items often resolve to a two-candidate choice near the end, which makes them prime flag targets. A flagging budget of 4–6 is realistic for most test-takers.
Verbal's average is closer to 1:57 per question, and reading-comprehension items share a passage — so revisiting one often means re-loading the passage mentally. Be conservative with flags in Verbal; 3–5 is plenty. Critical Reasoning questions where you narrowed to two choices are the best candidates.
Data Insights has the longest per-question average (~2:15) and the most complex prompts, including multi-part items like Two-Part Analysis and Table Analysis. Revisiting a DI question is expensive — you often have to rebuild the data interpretation from scratch. Cap flags at 3–4 and make sure the ones you use are genuinely two-candidate situations.
The fastest way to lock in the rules is to quiz yourself on the tricky edge cases. Each question below mirrors a real test-day decision.
Most of the points students lose around this feature aren't lost to the feature itself — they're lost to a small number of repeating misuse patterns. Spotting them ahead of time is mostly all the defense you need.
The single most common mistake is flagging 8, 9, or even 10 questions per section. When the review screen opens with a minute or two on the clock, the test-taker freezes — too many questions to revisit properly, not enough time to think, and only 3 edits to spend. The result is rushed changes made under stress, which is exactly how correct answers get swapped for incorrect ones.
The review screen invites second-guessing. An answer that felt right during the section can start to feel wrong once you're staring at it again. Unless you can verbalize a specific new reason to change it — a rule you remembered, a calculation you spot-checked, an eliminated option that now makes sense — don't edit. Gut-feeling edits swap right for wrong more often than the reverse.
Some students plan to rush the first few questions, pick anything, and "fix them later" with edits. This is the worst possible misuse. Early wrong answers lower the difficulty of subsequent questions, so even after your corrections the algorithm has already routed you to an easier pool with a lower score ceiling. You cannot out-clever the scoring system this way.
Worked Example
Setup: A student finishes Quant with 3:20 on the clock, having flagged 9 questions. On the review screen they start second-guessing answer after answer and end up editing three questions in the final 90 seconds — two of which they change from correct to incorrect.
| Mistake | Why It Hurts | What to Do Instead |
|---|---|---|
| Flagging 8+ questions per section | No realistic time to review them; decision fatigue at the end | Cap flags at the table budget and move on |
| Editing the same answer twice | Burns two of three allowed changes for the section | Commit to the new answer the first time or leave it alone |
| Editing on a gut feeling at the review screen | Swaps right answers for wrong ones more often than the reverse | Edit only when you have a specific new reason |
| Rushing early to 'correct later' | Early wrong answers lower later difficulty and cap your score | Play the first half clean; treat edits as back-half insurance |
| Ignoring the section clock while reviewing | Timer hits zero and the review screen closes | Leave about 2 minutes on the clock for the review screen |
| Not practicing the feature in mocks | Test-day is the first real attempt at using it well | Log flags and edits on every full-length practice test |
You will fall back on whatever habit you built in practice. If you never used the flagging and editing features on your mocks, you'll fumble them on test day — not because the feature is hard, but because it's unfamiliar.
Start with the official mba.com practice exams. They include the actual Review & Edit interface, bookmark icons, and timing behavior you'll see on test day. Third-party mocks may not replicate the exact UX of the review screen, so get at least a few reps on the official software before taking the real thing.
Decide your rule before you sit down: two-candidate flag only, three-to-six flags per section, edit only with a stated reason. Write it on your scratch sheet during the test-center setup window. A pre-committed rule removes decision fatigue in the moment, which is when most misuse happens.
After every full-length mock, log three numbers: flags placed, edits used, and edits that actually flipped a wrong answer to a right one. Over five or six mocks you'll see whether your editing is a net gain or a net loss. If you're editing four or five times across the whole mock but only turning one wrong answer into a right one, your flagging bar is too low.
The questions students ask most often about this feature tend to cluster around the same edge cases. Here are the direct answers.
You can edit up to 3 answers per section on the GMAT Focus Edition, for a maximum of 9 answer changes across the entire exam. The limit counts changes, not questions — if you edit the same answer twice, that consumes two of your three allowed changes for the section. The cap exists so the adaptive algorithm has enough fixed responses to score you accurately.
The Question Review & Edit screen appears automatically after you submit an answer to the last question in a section, but only if time still remains on the section clock. If you run out of time before reaching the final question — or immediately after answering it — the screen does not appear and you cannot review or change any answers in that section.
Bookmarking is an unlimited navigation aid: you can flag as many questions as you like while answering them so they are easy to find again. Editing is the act of actually changing an answer on the Review & Edit screen, and it is hard-capped at three per section. You are never required to edit a question just because you bookmarked it.