You took a practice test.
You waited for the score. You felt something — maybe excitement, maybe dread. You scrolled through the questions you missed. Maybe you read a few explanations.
Then you went back to studying the same way you were before. And a week or two later, you took another practice test.
If that sounds familiar, you're not alone. It's probably the most common pattern in GMAT® prep. And it makes complete sense — that's how every test you've ever taken has worked. Study the material, take the test, see how you did.
But the GMAT® doesn't work like other tests. And practice tests don't work the way most people think they do.
The Practice Test Is the Least Important Part of the Cycle
That sounds strange. But think about it like stepping on a scale when you're trying to lose weight.
Stepping on the scale is a critical step. You need to know where you are.
But stepping on the scale is not what makes you lose weight. Everything that happens between weigh-ins is what moves the number.
Practice tests work the same way. The test is a measurement. The growth happens between tests.
If you've been taking a practice test every weekend and your score isn't moving, that might be a big part of the problem. It's easy to step on the scale over and over, hoping for a different number, without doing the work between tests that would change it.
The Three-Layer Review
A lot of people skip the review, or do it superficially. They glance at the questions they got wrong, maybe read an explanation, and call it done.
A real review has three layers. And the order matters.
Layer 1: Timing
The first thing to look at after every practice test is not your score. It's your timing.
Here's why. The GMAT® scoring algorithm is not purely accuracy-based. Your score depends on the difficulty level of the questions you see, which is determined by the adaptive algorithm responding in real time to how you perform.
This means something counterintuitive: missing an easy question hurts your score more than missing a hard question. Getting a "gettable" question right helps your score more than getting a hard one right.
The most common reason people miss gettable questions is that they burned too much time on questions they were never going to get. They spent four minutes on a problem they had no strategy for, and then rushed through the last five questions of the section — which might include three or four they knew how to do.
So when you review timing, you're looking for one thing: did you have enough time to answer every question within your skill set?
If the answer is no, timing is your highest-priority fix. Not content. Not new strategies. Timing. Because if you're not managing your time well, you're not getting credit for what you already know. And in that situation, learning more usually won't move your score.
Layer 2: Two Separate Mistake Lists
Once your timing is solid, the next step is figuring out how many questions you actually knew how to do but got wrong anyway.
Split every incorrect response into two separate lists:
Category A — questions you knew how to do but got wrong. Misreads. Computation errors. Rushing through a step you understand. Trying to do too much in your head.
Category B — questions you didn't know how to do. Unfamiliar content. Strategies you haven't practiced. Question types you haven't seen before.
These are fundamentally different problems that require different fixes.
Category A is a habit problem. You need visual reminders, checklists, behavioral change. You need to build a habit of underlining the actual question being asked. More content knowledge won't fix a misread.
Category B is a knowledge problem. You need to re-learn the strategy, practice similar problems, build familiarity. A checklist about reading carefully won't fix a content gap.
Most people in the GMAT® community will tell you to keep one big "error log" of all your mistakes. We think lumping everything together makes it hard to see what's actually going on. It's like going to the doctor and saying "I feel bad" without distinguishing between a headache and a broken ankle. The diagnosis matters because the treatment is different.
Fixing Category A: Look at your full Category A list across your entire prep. Find the single most common cause of missed questions you knew how to do. Create one new rule to address it. Not "stop making computation mistakes" — something specific, like "check the previous step every time I write a new line on my scratch paper."
Then remind yourself of that rule often enough that forgetting would be harder than remembering. A note card on your desk. A daily email to yourself. A physical reminder you can't ignore.
Track your compliance. Did you execute the new behavior in today's study session? Three sessions in a row of perfect execution is the minimum before moving to the next habit.
The question isn't "how do I get a little better at this?" It's "what would it take to permanently delete this behavior?"
Fixing Category B: Take the first 20% of every study session and re-solve questions you didn't know how to do. Not new problems. The same ones. Over and over until the moment you read the question, you can see the path to the answer.
Re-solving old problems is not exciting. There's research on this — novelty is one of the top three pleasure-producing things for human beings, right up there with love and food. Our brains are wired to crave new things. New problems feel productive. Going back to old problems feels like admitting you didn't learn it the first time.
But what we see with students is consistent with the research on this. The key to building durable skills is a mix of old and new material, heavily weighted toward repetition. There's a concept in cognitive science called "desirable difficulty" — learning conditions that feel harder in the moment tend to produce better long-term retention. Re-solving a problem you struggled with last week feels like slogging. But it's building neural pathways that new problems can't build.
Layer 3: Question-by-Question Review
Now you can do what most people do first — go through each question one by one.
For correct answers, a quick check: did you use the best approach? For incorrect answers, sort them into Category A or Category B.
As you review, you're looking for patterns. A lot of what separates a 650 from a 700 is pattern recognition. At the 650 level, you might recognize problem types you've seen before about 40% of the time on a new test. At the 700 level, maybe 60%. You build that recognition not by blazing through thousands of new problems, but through targeted repetition on the problems that are challenging for you.
One thing that helps: write down your accuracy percentages by topic. Data sufficiency. Reading comprehension. Problem solving. Critical reasoning. Get the actual numbers.
Because your feelings about how you did are often different from how you actually did. There's a phenomenon in learning science called the illusion of competence, where people tend to be about three times more confident in their knowledge than they should be. When you think "yeah, I know rates problems, I'm good on those" — there's a decent chance your actual accuracy is lower than your gut says.
Get the numbers. Let the data show you where to focus.
What to Do Between Practice Tests
Pick a maximum of three focus areas. One per section if possible — one quant, one verbal, one DI.
Ask yourself: if you could wave a magic wand and be instantly better at one thing per section, what would turn the most wrong answers into right answers?
That's your focus list.
Roughly 80% of your study time between tests goes to those three areas. About 20% goes to maintenance — keeping everything else from getting rusty.
Go deep on three areas. Not "a little bit of everything." If you spread across eight topics, you'll see very little improvement in any of them. You'll take another practice test, see the same score, and feel like nothing's working.
But if you go deep on three — working through every official guide question on that topic, building your accuracy to 80% or higher at an appropriate pace — you'll walk into your next practice test with improved skills in the areas that matter most.
When to Take Your Next Practice Test
Take your next practice test when your skills are ready. Not when the calendar says it's time.
If you've been taking a test every weekend without improving between tests, the test frequency is likely part of the problem. Each practice test generates a list of focus areas. Those focus areas take time to address. Two weeks. Four weeks. Maybe six weeks if you're studying a few hours a week.
If you take another test before you've done the work, it's probably going to tell you roughly the same thing as last time. And you'll have spent three hours on a Saturday that could have gone toward building skills.
The right time for your next test: when you've resolved your top three focus areas to 80% accuracy at an appropriate pace. When you've re-solved your Category B questions and can get them right consistently. When you've built new habits to reduce Category A errors.
That's when the next test gives you useful, new information.
The Emotional Side
A lot of what we've described is logical. But when you finish a three-hour practice test and see a score that's 50 points below what you were hoping for, logic is not what's running the show.
Twenty years of conditioning says a test score measures how smart you are. Even if you consciously know it's just a data point, your nervous system might be telling you something different.
If you've felt that way, you're normal. The emotional reaction is real.
Here's a framework that helps. Any time you catch yourself thinking about the result — the score, whether you're going to hit your goal — that's outcome thinking. And outcome thinking tends to work against performance.
What you want instead is process thinking. Am I writing things down? Am I answering the right question? Do I have a strategy for this problem? Am I managing my time well?
Use outcome thinking as a trigger. Every time you catch yourself thinking about the score, redirect to process.
Feel whatever you feel after a practice test. Give yourself 24 hours if you need to. Then come back to the review process. The score is just telling you where to focus next. It's not a referendum on your intelligence or your worth. It's a data point.
The Full Cycle
The practice test cycle has three phases:
Phase 1: The test. Simulate real conditions. Proper environment, proper timing, no phone. The data is only as useful as the conditions are realistic.
Phase 2: The review. Layer 1: timing analysis. Layer 2: Category A vs Category B error separation. Layer 3: question-by-question review. This should take at least as long as the test itself.
Phase 3: Targeted study. Two to six weeks of focused work. 80% on your top three areas. 20% on maintenance. First 20% of every session goes to re-solving Category B questions.
Then repeat. Test, review, targeted study. If your cycle is working, you'll usually see something move. And if nothing moves, that's not a failure — it's data. It means one of the three layers needs more attention.
FAQ
How often should I take GMAT® practice tests?
For most students, one practice test every two to six weeks is ideal. The test is a measurement tool, not a learning tool. The growth happens between tests. Take your next practice test when you've improved your focus areas — not when the calendar says it's time.
What is a Category A mistake on the GMAT®?
A Category A mistake is a question you knew how to do but got wrong anyway. Misreads, computation errors, rushing through a step you understand. Category A mistakes are different from Category B mistakes (questions you didn't know how to do) because they require habit changes, not content review.
Should I keep an error log for GMAT® practice?
A single error log that lumps all mistakes together is rarely effective. Instead, split your mistakes into two separate lists: Category A (execution errors) and Category B (knowledge gaps). These require different fixes, so combining them makes it hard to see what's going on.
How long should I spend reviewing a GMAT® practice test?
Your review should take at least as long as the test itself. If you spent three hours taking the test and twenty minutes reviewing it, the ratio is the opposite of what you want. A proper review includes timing analysis, error categorization, and question-by-question review.
What should I study between GMAT® practice tests?
Pick a maximum of three focus areas. Spend about 20% of each study session re-solving questions you got wrong previously, and divide the rest across your focus areas and maintenance. Don't take another practice test until you've measurably improved in those three areas.
Want to learn even more?
We walk through this entire system in detail in Episode 48 of our podcast series, "How to Review GMAT® Practice Tests." Isaac covers the three-layer review framework, how to build habits that permanently eliminate Category A mistakes, and how to handle the emotional weight of practice test scores.