If you know the math but can't finish the section, you're in the right place.
Most GMAT® Quant advice tells you to work on your mental math. Learn to calculate faster in your head. Do more practice problems. Push yourself to go quicker.
That advice is wrong for almost everyone. And following it can make you slower, because mental math tends to lead to mistakes, and mistakes cost more time than writing things out.
Speed on GMAT® Quant doesn't come from trying to go fast. It comes from process. When your scratch work is organized, your arithmetic is automatic, and your pattern recognition is sharp, you move through questions quickly without forcing it. The speed is a byproduct of doing everything else well.
Why You Might Be Slow
If you're running out of time on GMAT® Quant, the problem probably isn't that you need to calculate faster. The time pressure on this section can feel relentless, and if you've been practicing for a while and still can't finish, it's easy to start questioning whether something is wrong with you. There isn't. The bottleneck is almost always process, not ability.
It's more likely one of four things:
You're doing too much math in your head. Mental math feels faster because you're not writing anything down, but it can cost you on the back end when you have to double-check your work or redo a step because you lost track of a number. You might multiply 45 by 12 in your head, get 540 instead of the correct 504, and two steps later your answer doesn't match any choice. Now you're retracing every step to find where things went wrong. Writing things out takes a few extra seconds per step, but it cuts out most of the rework that eats minutes.
You're choosing the wrong approach. When you don't have strong pattern recognition, you start each question from scratch. You try algebra, it doesn't work, you try picking numbers, that doesn't work either, and now you've spent two minutes without a clear path. A student with better pattern recognition sees the problem type in the first five seconds and knows the approach before they start writing.
You're not letting go. You spend four minutes on a hard question hoping a strategy will appear, then have to rush through the last five questions. The time you invested in that one question cost you accuracy on three others.
Each of these is a system problem, not a knowledge problem. You don't need to learn more math. You need better systems for executing the math you already know.
Lever 1: Write Everything Out
This is the most counterintuitive speed tip we give, and it's also the most important.
The GMAT® Quant section doesn't give you a calculator. Every arithmetic operation, from adding two-digit numbers to dividing decimals, has to be done by hand. A lot of instructors will tell you to work on your mental math skills to save time.
In our experience, that advice is wrong for almost everyone. Mental math can lead to computation errors, and computation errors tend to lead to rework. When you do a calculation in your head and get it wrong, you don't always catch the mistake immediately. Sometimes you catch it two steps later, when your answer doesn't match any of the answer choices. Now you have to go back, find the error, and redo the work. That's two or three minutes lost on a question you should have finished in 90 seconds.
Writing everything out takes slightly longer per step, but it eliminates almost all rework. You can see your work on the page, check each step as you go, and catch errors before they cascade.
The key is writing it out in an organized way, not scattering numbers across the scratch pad. A labeled, step-by-step solution on your scratch pad is faster than a disorganized one, because you can find what you need when you need it.
If you've been doing math in your head for years, this will feel slow at first. That's normal. The speed comes later, once the habit is established and you stop losing time to errors and rework. Train yourself to write out every step, even when the calculation seems simple, even when you're confident you can do it mentally. The goal is to make organized scratch work your default, so it happens automatically under pressure.
Lever 2: Build Arithmetic Fluency
GMAT® Quant tests math through roughly high school level. The concepts aren't advanced, but the arithmetic can be tedious. If you haven't done long division by hand in ten years, that's a skill gap that directly affects your speed.
The good news is that arithmetic fluency is the easiest lever to build. It's pure repetition. You don't need to understand new concepts. You just need to relearn the procedures and do them enough times that they become automatic.
Start with the operations that show up most frequently on the GMAT®:
Fraction-Decimal-Percent Conversions
You should know the common conversions cold: , , , , , , etc. These conversions show up in word problems, percent problems, and data sufficiency. If you have to derive them each time, you're losing 10 to 15 seconds per occurrence.
Long Division and Long Multiplication With Decimals
These are procedures you learned in elementary school and probably haven't used since. They're not hard, but they need to be fast and automatic. Practice dividing two-digit numbers into three-digit numbers, multiplying decimals, and converting remainders to fractions.
Squares and Cubes
Know through and through without thinking. These show up in number properties, exponents, and ratio problems. If you can recognize that instantly, you save time on factor problems, prime factorization, and square root estimation.
Prime Numbers Up to 30
Know which numbers are prime: 2, 3, 5, 7, 11, 13, 17, 19, 23, 29. Prime factorization is one of the most common operations on the GMAT®, and being able to instantly recognize primes saves you from testing divisibility one number at a time.
You can build all of this with flashcards or a simple drill app. Ten minutes a day for two weeks should make a measurable difference. The goal is to make the common operations automatic, so your working memory is free for the actual problem-solving.
Lever 3: Develop Pattern Recognition
This is where the biggest speed gains come from, and it's also the lever that takes the longest to build.
Pattern recognition is the ability to look at a GMAT® question and immediately know what type of problem it is and what approach will work. When you have strong pattern recognition, you skip the trial-and-error phase that eats time. You read the question, recognize the structure, and start executing the right approach immediately.
A student with weak pattern recognition reads a mixture problem and thinks, "Okay, what do I do here? Maybe I set up an equation? Maybe I pick numbers? Maybe I use the answer choices?" That's two minutes of deliberation before any math happens.
A student with strong pattern recognition reads the same problem and thinks, "Mixture problem. The amounts combine but the concentration changes. I'll set up a table with the two solutions and the combined mixture." They're writing before the first minute is up.
Pattern recognition comes from volume, from seeing enough problems of each type that the structural cues become familiar and the approach surfaces before you've consciously decided what to do. But there's a catch: brute-force volume usually doesn't work on its own. Doing 500 random practice problems without reviewing them doesn't build pattern recognition. It builds exposure, which is a different thing.
The right way to build pattern recognition is to do problems in topic-specific batches, review them thoroughly, and then re-solve the ones that took you too long or that you got wrong. The re-solving is what builds the pattern. When you re-solve a problem, you're not just practicing the math. You're practicing the recognition: "Oh, this is what a mixture problem looks like. This is the cue. This is the approach."
Try this approach over six to eight weeks:
- Pick one topic (mixture problems, rate problems, number properties, etc.)
- Do 10 to 15 problems from that topic over two days
- Review every problem afterward. For each one, note: what type is it? What was the cue? What approach worked?
- Re-solve the ones that took more than two minutes or that you got wrong. Do them again the next day.
- Move to the next topic and repeat.
After six to eight weeks of this, you'll have seen enough variations of each problem type that the recognition starts happening automatically. You'll read a question and the approach will come to you without deliberate effort. That's when you start getting faster.
Using official materials matters here. Third-party questions can look similar to GMAT® questions, but the structural cues are slightly different. If you build your pattern recognition on non-official questions, you may not transfer that recognition to the real exam. Use the Official Guide and mba.com practice materials as your primary source.
Lever 4: Train With a Count-Up Timer
Most people train with a countdown timer because that's what the real exam uses. But a countdown timer creates pressure, and pressure makes you rush, and rushing makes you skip steps, and skipping steps makes you slow (see Lever 1).
A count-up timer does the opposite. It removes pressure entirely. You can take as long as you need on each question, and the timer just tracks how long you took. The goal isn't to finish quickly. The goal is to execute perfect process on every question, and let the speed develop on its own.
The phases:
Phase one (no time pressure): Set a count-up timer and work through 10 questions. Take as long as you need. Focus on organized scratch work, clean arithmetic, and choosing the right approach. Don't worry about how long each question takes. When you finish, note your times. Some questions may take four or five minutes. That's fine.
Phase two (awareness): After a week or two, look at your times. You'll notice that some question types naturally take less time than others. Your pattern recognition is already starting to work. Keep using the count-up timer, but start paying attention to which questions are slow and why. Is it the approach? The arithmetic? The reading?
Phase three (gentle pressure): Once your scratch work is consistently organized and your approach selection is solid, add a gentle countdown. Try 10 questions in 25 minutes instead of the 45 you'd get on the real exam. This gives you 2.5 minutes per question instead of the usual 2.1, so there's pressure but it's not crushing.
Phase four (real conditions): Only move to full exam timing when you're consistently finishing 10-question sets under 25 minutes. At that point, switch to the real 45-minute, 21-question format and see where you stand.
The mistake most people make is jumping straight to phase four. They take full-length practice exams before they've built the underlying skills, and the time pressure forces them into bad habits: mental math, skipped steps, disorganized scratch work. Then they review the test, see that they ran out of time, and conclude that they need to go faster. So they do more practice exams, and the cycle repeats.
If this sounds familiar, try going back to phase one. Build the process first. The speed should follow.
Common Speed Traps
A few specific habits tend to slow people down on GMAT® Quant. If you're running out of time, check whether any of these show up in your practice.
Refusing to Let Go of Hard Questions
Some questions are designed to eat your time. If you've read a question twice and still don't have a clear approach after two minutes, guess and move on. The two minutes you spend guessing is much less expensive than the four minutes you'd spend struggling, plus the rushed work on the next three questions. We have a full guide to when to let go of a GMAT® question that covers this in detail.
Testing Numbers Without a Plan
Picking numbers is a valid strategy on many GMAT® questions, but it can be slow if you don't have a system for choosing good numbers. If you test three or four sets of numbers before finding one that eliminates the wrong answers, you've spent three minutes on a question that should have taken 90 seconds. The fix is to learn the standard test cases for each problem type (0, 1, -1, fractions, and the boundary values of the constraint) so you test efficiently.
Doing Algebra When Picking Numbers Would Be Faster
Some GMAT® questions are designed to make algebra messy. The equation looks solvable, but the manipulation takes four steps and creates opportunities for errors. For these questions, picking numbers from the answer choices can be faster and more reliable. The key is recognizing which questions favor which approach, which comes back to pattern recognition.
FAQ
How long does it take to get faster at GMAT® Quant?
Most students see noticeable speed improvements after four to six weeks of focused practice. Arithmetic fluency can improve in one to two weeks with daily drills. Pattern recognition takes longer, usually six to eight weeks of topic-specific batching and review. If you've been stuck at the same speed for more than two months despite consistent practice, you may need to change your approach rather than do more of the same.
Should I use mental math to save time on GMAT® Quant?
For almost everyone, no. Mental math feels faster but can lead to computation errors that cost more time on the back end. Writing out your work takes slightly longer per step, but it tends to eliminate rework and helps you catch mistakes before they cascade. There are a few people who are naturally fast and accurate with mental math, and if that's you, keep doing what works. But for most test-takers, organized scratch work is faster than mental math.
How many practice problems should I do per day to build speed?
Quality matters more than quantity. Ten problems with thorough review and re-solving is more valuable than 30 problems done quickly and never reviewed. If you're doing topic-specific batches (which we recommend for building pattern recognition), 10 to 15 problems per day with full review is a good target. The review process, not the problem count, is what builds speed.
What if I'm already good at the math but still running out of time?
If you know the concepts but can't finish, the problem is probably process, not knowledge. Check your scratch work organization, your question-reading habits, and your ability to let go of hard questions. Also check whether you're using the right approach for each question type. Strong math skills with weak pattern recognition will leave you spending too long on questions you should be able to answer quickly.
Should I drill with a timer?
Start with a count-up timer and no time pressure. Build organized scratch work and clean process first. Once that's consistent, add gentle time pressure (2.5 minutes per question). Only move to full exam timing when you're finishing sets comfortably under the gentle target. Jumping straight to full timing before your process is solid will reinforce bad habits.
Will learning mental math tricks help on the GMAT®?
Most mental math tricks are designed for specific calculation types (multiplying by 11, squaring numbers ending in 5, etc.) and probably don't come up often enough on the GMAT® to justify the learning time. The arithmetic that matters on the GMAT® is more basic: fraction-decimal conversions, long division, long multiplication, and recognizing common squares and primes. Build fluency with those operations rather than learning party tricks.
Want to learn even more?
- GMAT® Timing Strategy
- How to Build GMAT® Quant Confidence When You Feel Bad at Math
- How to Study for the GMAT®: A Complete Guide
- How to Break Through a GMAT® Score Plateau
You can also listen to our podcast on Spotify, Apple Podcasts, or YouTube for more strategy episodes and worked solutions.