How to keep a GMAT error log that actually raises your score
Two people can do the same 500 practice questions and end up with very different scores. The difference is rarely how many questions they did. It is what they did after getting one wrong. One checks the answer, reads the explanation, nods and moves on. The other writes down why the mistake happened, what to do differently, and comes back to the question a few days later to prove the fix worked.
An error log is the tool for the second approach. Done badly, it is a tedious spreadsheet that nobody reads. Done well, it is the most accurate map you will ever have of what stands between you and your target score. This guide explains what to record, how to classify mistakes so patterns appear, a free template you can copy, and the review routine that turns the log into points.
Why an error log works
Every miss contains two pieces of information: the question you got wrong and the reason you got it wrong. Most test-takers keep only the first. The reason is the useful part, because reasons repeat. If you miss ten questions for ten different reasons, you have ten small problems. More often, you miss ten questions for two or three reasons, and fixing those reasons fixes future questions you have not seen yet.
The log also protects you from a comfortable illusion. Reading an explanation feels like learning, because everything makes sense while you read it. The real test is whether you can solve the same question cold a few days later. Scheduled redos, where you try the question again without looking at the solution, are what turn "that makes sense" into "I can do that".
What to log (it is not only wrong answers)
Log three kinds of question:
- Wrong answers. The obvious ones.
- Lucky guesses. If you got it right but were choosing between two answers, you do not reliably know it. Treat it as a miss.
- Right but slow. If a question took well over the average (about two minutes on this exam), log it. Slow correct answers cost points indirectly, because they take time from later questions and force guesses at the end of a section.
Do not log questions you got right quickly and confidently. The log should be short enough that you will actually review it.
The root-cause categories
The most important column is root cause. Use a small, fixed set of categories so you can count them. These six cover almost every GMAT mistake:
| Category | What it means | Typical fix |
|---|---|---|
| Content gap | You did not know the rule, formula or concept | Learn it, then do 5–10 targeted questions on that topic |
| Misread | You answered a different question from the one asked, or missed a word such as "not", "integer" or "distinct" | Restate the question in your own words before solving |
| Process | You knew the content but chose a slow or wrong approach | Learn the better method; note the signal that should trigger it |
| Execution | Arithmetic slip, sign error, copied a number wrong | Write cleaner scratch work; check the step you most often get wrong |
| Trap answer | You fell for a designed wrong answer (in Verbal: out of scope, too extreme, wrong direction) | Name the trap family; eliminate by family next time |
| Timing | You ran out of time or rushed because earlier questions ran long | Set a per-question ceiling and practise guessing on purpose |
Be honest with yourself here. "Careless" is the most tempting label and the least useful one. Most so-called careless errors are either misreads (a process problem you can fix with a routine) or execution errors in one recurring step (a specific skill you can drill).
A free GMAT error log template
Copy these columns into a spreadsheet, a notebook or a notes app. One row per logged question.
| Column | What to write | Example |
|---|---|---|
| Date | When you attempted it | 2026-09-23 |
| Source | Where the question is from, so you can find it again | Practice set, question ID or page |
| Section and type | Quant PS, Verbal CR or RC, Data Insights DS, TPA, TA, GI or MSR | Data Insights, DS |
| Topic | The narrowest useful label | Number properties: even/odd |
| Result | Wrong, lucky guess, or slow | Wrong |
| Time | How long you spent | 3:10 |
| Root cause | One of the six categories | Process |
| What happened | One sentence, specific | Used statement (1) info when judging (2) |
| Rule for next time | One sentence you could say to yourself on test day | Judge (2) as if (1) did not exist; cover it with my hand |
| Redo 1 (about 3 days) | Date and result | 2026-09-26: right, 1:40 |
| Redo 2 (about 10 days) | Date and result | 2026-10-03: right, 1:25 |
Two columns do most of the work. "What happened" forces you to diagnose instead of summarise, so "got it wrong" is not allowed. "Rule for next time" turns the diagnosis into a behaviour. If you cannot write a rule, you have not understood the mistake yet; reread the explanation until you can.
A question leaves the log once you have got it right, at a reasonable pace, on two separate redos.
Writing entries that are actually useful
Compare two entries for the same Critical Reasoning miss:
Weak: "Chose C, answer was B. Didn't read carefully."
Useful: "Weaken question. Picked C because it was about the plan's cost, but the conclusion was only about whether the plan would reduce waiting times. Rule: before the choices, write the conclusion in five words and check every choice against it."
The useful entry names the question type, the exact trap (out of scope) and a rule you can apply to a new question. Writing it takes a minute longer. It is also the only one of the two that can change what you do next time.
For Quant and Data Sufficiency, the same idea applies. "Arithmetic mistake" is weak. "Divided by 0.8 instead of multiplying when undoing a 20% discount" is useful, because it tells you which step to check from now on.
The redo schedule
Redo each logged question twice, spaced apart, without looking at your notes first:
- Redo 1 after about 3 days. Long enough that you cannot remember the answer, short enough that the lesson is still fresh.
- Redo 2 after about 10 days. This checks whether the fix has stuck.
If you miss a redo, reset the question: write down why you missed it this time (often a different reason from the first) and schedule it again. Be honest about "remembered the answer" versus "solved it". If you remember the answer, solve it anyway and check whether your method was sound.
The weekly review
Once a week, set aside 30 to 45 minutes to look at the log as a whole, not question by question:
- Count by root cause. Which category has the most entries this week? That is your process fix for next week.
- Count by section, type and topic. Which two or three topics recur? Those get targeted practice next week.
- Read your rules. Rules that appear more than once are your personal test-day checklist.
- Check timing. If "slow" entries cluster in one question type, that type needs timed practice, not more untimed work.
Then adjust your next week's plan: roughly half your practice time on the top two or three problem topics. Our realistic GMAT Focus study plan shows how an error log fits into a 4, 8 or 12 week schedule.
Mistakes people make with error logs
- Logging everything. A log with 400 entries and no review is a diary. Log misses, lucky guesses and slow answers only, and keep the weekly review.
- Copying the explanation. The explanation is already written. The log is for your reason and your rule.
- Too many categories. Twenty categories means no category ever has enough entries to show a pattern. Stick to six.
- Never redoing questions. Without redos, you never find out whether a fix worked.
- Giving up in week three. The value of a log grows over time, because patterns only show up across dozens of entries. A short log you keep beats a detailed one you abandon.
Letting software keep the log
Maintaining a spreadsheet by hand is the part most people drop. GMAT Drill records every question you miss automatically and lists it with its date, question type, topic and difficulty, so the list of misses for your weekly review builds itself. You still need to write the reason and the rule yourself, in your own notes or spreadsheet, because nobody else can diagnose your thinking. Pair it with the free practice sets for Data Sufficiency, Problem Solving and Critical Reasoning, and use the GMAT Focus score calculator to see what a gain in your weakest section would do for your Total Score.
If your misses cluster in particular question types, our guides to Data Insights question types and Critical Reasoning question types go through the method and the usual traps for each.
Frequently asked questions
What should a GMAT error log include?
At minimum: the question source, section and question type, topic, whether it was wrong, a lucky guess or slow, the time taken, a root-cause category, one sentence on what happened, one rule for next time, and the dates and results of your redos.
Should I log questions I got right?
Log them if you guessed between two choices or took well over two minutes. Both are signs you do not reliably know how to solve that kind of question. Do not log fast, confident correct answers.
How often should I review my GMAT error log?
Redo individual questions about 3 days and about 10 days after the miss, and review the whole log once a week to count mistakes by root cause and topic. Use the weekly count to decide what to practise next.
Is a spreadsheet or a notebook better for an error log?
Either works. A spreadsheet makes it easy to filter and count by category, which helps the weekly review. A notebook can be quicker to fill in. The best format is the one you will keep using for the whole of your preparation.
Can an error log help with timing problems?
Yes. Logging slow correct answers, not just wrong ones, shows which question types eat your time. Those types then get timed practice and a personal time limit, so you learn when to guess and move on.