GMAT® Math Problems with Solutions: How to Use Them to Raise Your Score

Looking for GMAT® math problems with solutions is one of the more sensible instincts in test prep. You're probably not hunting for confirmation that the answer is (B) so much as for how the thinking goes, and watching a problem get worked step by step is how most of us learned math the first time around. Going back to that well makes sense.

There's a gap between reading solutions and studying from them, though, and the gap is wider than it looks. A solution read passively builds familiarity, and familiarity feels a lot like knowing. Familiarity alone doesn't score points on this exam, though, because the points come from what you can produce on a blank scratch pad, with a pen, while a clock runs. The fix isn't to stop reading solutions. It's an order of operations: attempt the problem, commit to an answer, read the solution as a comparison, and schedule the return trip.

Why worked solutions teach, and where they stop

There's real research behind the instinct. Learning science has studied worked examples for a long time, and the finding that keeps coming back is that a fully worked example lowers the mental load while a new process goes in. A worked example means you don't have to reconstruct every step from nothing. You watch a clean execution of the process instead, which leaves room to notice why each step happens.

The stop is what comes next. A solved problem can show you a process, but the exam asks you to produce one: from a blank pad, in about two minutes, without being told in advance which approach the problem needs. Producing and recognizing are different skills. A study plan built mostly on reading solutions trains the second skill and calls it preparation for the first, and that mismatch tends to surface on test day, where it's expensive.

The attempt comes first

The order that holds up in practice:

• Attempt the problem completely before reading anything. That means scratch work on the pad and an answer selected, the same way the real exam would run it. This is the step that turns a solution into feedback, because it gives the solution something to compare against.
• Commit to an answer, even when the commit is a guess. Where your judgment pointed when your knowledge ran out is often the most instructive place the solution can reach.
• Read the solution as a comparison of processes, not an explanation to memorize. If your path and the written path diverge, that junction is the lesson: a cleaner route to the same answer, or a different route because yours carried a hidden step.
• Add the problem to a redo list when it earns the spot. The criterion isn't only wrong answers. A question that took longer than you wanted, or where you had no strategy when you started, goes on the same list, with a one-line note about what you want to remember next time you see it.

Re-solving is what turns the list into a study tool. Three problems from the list, re-solved from a blank pad every study session, is the starting point we recommend, and five works when time allows. One ceiling matters, though: re-solving is review, and as a working guideline it should stay under about 20% of your total study time, because past that point it starts crowding out the fresh problems that add new coverage.

Keep re-solving separate from speed training. Speed work runs on fresh problems, because a problem you've already seen tests recognition, and recognition tends to make your times look better than they are. The redo list exists for accuracy and process, and our speed-building guide keeps the two in separate lanes for that reason.

What a good attempt looks like

An attempt needs the same conditions as the real thing, or the comparison afterward teaches the wrong lesson.

Give each problem its own space on the pad, write every step in one line, and keep the arithmetic out of your head. A large part of the attempt's value is the scratch work record, because it shows where the process wobbled, not only where it broke. Our guide to how writing down more improves your score covers what goes in each section of the pad, and the scratch pad setup guide covers the physical organization.

If a timer helps, time the attempt without letting it steer. The goal at this stage is an honest execution, not a fast one. Speed has its own program, and mixing the two is how sloppy process gets rehearsed.

The method on a real problem

Try this one before reading on.

The efficient path runs through percents of the right base. Retail is 20% above the wholesale cost, so retail is $1.2 \times 200 = 240$. The employee discount comes off retail, so the employee pays $0.9 \times 240 = 216$.

The same path in fractions, if you prefer them: 20% above cost is $\frac{6}{5}$ of cost, and 10% off is $\frac{9}{10}$ of retail.

$\frac{6}{5} \times \frac{9}{10} \times 200 = 216$

Answer choice (C), $220, is built from a specific slip: subtracting 10% of the wholesale cost (which is $20) instead of 10% of the retail price. If you chose (C), the redo note is specific: the discount applies to the latest price, not the original one. If your path matched the efficient one, compare the two approaches briefly and move on. Either way, the review connects to a decision you made during the attempt, so the next re-solve has a real question to answer.

The full walkthrough of this problem, with a column setup and both paths side by side, is in our worked solution for it.

Where the problems should come from

Official sources should lead, because the writing style is one of the things you're practicing. Third-party problems can be useful for drilling a particular skill, but they tend to differ from real GMAT® questions in wording, trap construction, and the reasoning the question rewards. Training mostly on approximations can leave a gap for test day, and closing that gap costs little, because official material is plentiful:

• The two free official practice exams in the MBA.com Starter Kit, each usable twice
• The Official Guide, current edition, with an explanation for every question (older editions work too; skip the question types the Focus Edition dropped)
• GMAC's official online question banks when you finish the book

Our guide to using the Official Guide effectively covers the sequencing, and if you're weighing whether the OG alone can carry a whole prep, we take up that question in a separate post.

A starting library, by topic

The worked solutions on this blog follow a standard format: the problem up front, a prompt to try it first, then the thinking written out. For rates and work, start with one machine 4 hours and another 3 and pumps A, B, and C running together, then add Juan's 11-seconds-per-yard setup, which brings unit conversion into the rate chart. On percents, the newspaper revenue problem puts variables in every answer choice, which makes it good practice for committing to a method instead of hunting for the answer. The lowest integer divisible by everything from 1 through 7 covers divisibility from the prime-factor side, and on the algebra side, swapping x for 1/x inside a squared fraction and absolute values with xy = 0 are the two we'd hand a student first. The weightlifting competition total and a decimal sequence with exponents round out statistics and sequences.

Where this study method breaks down

Study plans built around solved problems tend to run into three failure modes, and none of them come from a lack of effort.

Reading solutions back to back. An hour of solution-reading produces a smooth, familiar feeling, and that feeling doesn't sort processes you own from processes you watched. The attempt-first order separates the two, because feedback only exists when something was risked first.

Peeking mid-attempt. The moment a hint arrives, the problem converts from a solve into a read, and the redo list loses its most honest entry. If you're stuck, commit to your best guess and read on. A list note that says "no strategy for this setup" is still a real note.

Keeping the list and never returning to it. The classic error-log pattern, where the list grows and the return trips never happen, hasn't produced good results that we've seen. Attaching re-solving to the session structure, three per session whatever the list looks like, is what should keep the loop closed. Our error log guide covers the two-list version, including what belongs on each list.

There's a lived version of this, too. One member of this team put thousands of hours into GMAT® prep years ago and watched the score go down anyway, and the review loop described in this post is what got rebuilt from that experience. The list of mistakes was organized the whole time. The return trips were what was missing.

FAQ

Where can I find GMAT® math problems with solutions?

Start with official sources: the two free practice exams in the MBA.com Starter Kit, the Official Guide with its explanation for every question, and GMAC's official online question banks for volume beyond the book. This blog also keeps a library of worked solutions organized by topic.

Is reading solutions a good way to study for the GMAT®?

As a first exposure to a problem type, yes. As the main activity, it tends to build recognition rather than recall, and the exam pays for recall. Attempt each problem fully before reading its solution, then compare processes and send the problem to a redo list if it took too long or had no strategy.

How many GMAT® problems should I re-solve per session?

Three per study session, re-solved from a blank pad, is the starting point. Up to five works if time allows. Keep re-solving under about 20% of total study time so review doesn't crowd out fresh problems.

Should I re-solve problems I answered correctly?

If the answer was right but slow, or the strategy was improvised, yes. The redo list criterion is time and strategy, not just correctness, and a slow right answer is exactly the kind of execution worth automating.

Are third-party GMAT® practice problems okay to use?

They can help for drilling a specific skill, but official problems should lead. Third-party wording, trap patterns, and reasoning logic tend to differ from the real exam's, and the question style is part of what you're practicing.

How long should I wait before re-solving a problem?

Long enough that you're solving it from scratch rather than remembering the previous solution. A day or two is usually enough; if the answer shows up from the first line of scratch work, give it more time.

Want to learn even more?

Related reading:

• How to Review GMAT® Practice Tests: A Three-Layer System
• GMAT® Error Log: How to Track and Review Mistakes
• How to Use the GMAT® Official Guide Effectively
• Is the GMAT® Official Guide Enough to Prepare for the Exam?
• GMAT® Quant Preparation: How to Study for the Quant Section
• How to Get Faster at GMAT® Quant: A Practical Speed-Building Guide
• Be Accurate to Improve Your GMAT® Score: Why Precision Comes Before Speed
• Best GMAT® Study Materials and Resources for 2026

For worked problems with the thinking out loud, the Real GMAT® Problems episodes of our podcast are the format in its natural habitat. The work-rate episode page is a good place to start, and you can find the show on Spotify, Apple Podcasts, or YouTube.