Arithmetic on the GMAT Focus is the backbone of the Quant section. Even with only 31 questions in the whole section, the arithmetic cluster — number properties, fractions and decimals, percentages, ratio and proportion, averages, and the rate–time–distance family — accounts for the majority of stems a candidate will see on test day. A learner who treats arithmetic as 'the easy stuff' tends to leak points exactly where the section's adaptive algorithm punishes hesitation. The clusters below are the ones I drill with every candidate in the first three weeks of preparation, because the score movement from 78 to 83 lives or dies in this material.
What 'arithmetic' actually means on the GMAT Focus Quant section
On the GMAT Focus, arithmetic is not a separate subsection. It is a thread that runs through Data Sufficiency and Problem Solving items, often invisibly, behind what looks like a word problem or a geometry diagram. The test is designed so that roughly half the 31 Quant questions test your ability to manipulate numbers under time pressure, regardless of whether the stem announces itself as a number-property puzzle, a percentage problem, or a word problem about work rates. A candidate who walks in expecting 'arithmetic' to mean long division will be ambushed by stems that disguise the same operations inside layered language.
Two structural points shape how the section handles arithmetic. First, the GMAT Focus uses a multi-stage adaptive format: the second module of Quant branches harder or easier based on performance in the first. Getting the arithmetic wrong in the first 6–7 questions does not just lose marks — it actively removes the harder arithmetic items from your question pool, which compresses the score ceiling. Second, the section is unanswerable without a working command of the underlying arithmetic concepts rather than a calculator-driven shortcut habit. The on-screen calculator is permitted, but reaching for it on every item costs more time than the calculation itself. Mental estimation, factor recognition, and the ability to simplify before multiplying are the real currency of the section.
In practice, the arithmetic items on the GMAT Focus break into six recognisable families. Each family has its own stem vocabulary, its own typical trap structure, and its own pacing budget. Most candidates in my caseload have two families that are genuinely weak and four that feel fine until the test exposes the gaps. Diagnosing which families are weak is a faster path to a 5-point gain than grinding generic mixed sets, and it is the first move I make in any preparation plan.
Six arithmetic question families you will see on the GMAT Focus
The six families below cover the arithmetic stems that appear most often. They are not the official syllabus — the test does not publish one — but they map cleanly to the item types that experienced tutors recognise across official and third-party materials.
- Number properties and divisibility: primes, factors, remainders, digit sums, parity, LCM and GCD reasoning.
- Fractions, decimals, and percentage conversions: shifting fluently between 1/8 = 0.125 = 12.5%, and reverse-engineering a percentage to find the original whole.
- Ratio, proportion, and mixture problems: splitting a quantity, scaling a recipe, combining solutions of different concentrations.
- Averages, weighted averages, and alligation: mean as balance point, weighted averages when groups have unequal size, the alligation shortcut for mixture problems.
- Rate–time–distance and work-rate problems: D = R × T in motion contexts and its analogue W = R × T for combined worker problems.
- Percent change, successive percentages, and growth/decline: applying a percentage increase or decrease more than once, and recognising that two successive 10% moves are not 20%.
Each family can appear as a stand-alone Problem Solving stem or as one of the two statements in a Data Sufficiency question. Data Sufficiency forces an extra layer of judgement: rather than computing an answer, you decide whether the given statements are sufficient, sometimes in combination, sometimes alone, sometimes not at all. Many candidates underestimate how much Data Sufficiency depends on the same arithmetic fluency. If you cannot tell whether a remainder will always be even, you cannot rule out a Data Sufficiency statement, and the question is lost.
Number properties: the family most candidates underestimate
Number properties is the family where test-takers who are 'good at maths' lose the most unexpected points. The questions look childish: is n divisible by 6, is the product of two consecutive integers even, what is the units digit of 7^14. The challenge is rarely the underlying concept — it is the discipline of working through all possible cases, including the ones the stem does not flag. A question that says n is a positive integer will sometimes allow n = 1; a question that says n is a two-digit number will sometimes need you to consider the lower and upper bound together.
Three sub-skills carry most of the weight. First, prime factorisation fluency: any integer can be broken into primes, and divisibility rules are consequences of which primes appear in that factorisation. Second, parity reasoning: a sum is even when both terms are even or both are odd, a product is odd only when all factors are odd, and these rules survive any number of layers. Third, remainder arithmetic: if n leaves a remainder of 3 when divided by 7, then n + 4 leaves a remainder of 0, and this kind of shift is the engine of a large class of Data Sufficiency items.
On the GMAT Focus, number-property questions also appear as disguised form. A stem about a lock with a three-digit code, or about the order in which runners cross a finish line, often has a number-property question hiding inside it. The test is paid to reward flexible thinking, not to label the family for you. The candidates I work with who score above 82 in Quant all share one habit: when they read a number-property stem, they write down the constraint and at least one boundary case before they look at the answer choices. That ten-second pause is the cheapest insurance on the section.
Fractions, decimals, and percentages: the conversion trap
The conversion trap is the single most common arithmetic error I see. A stem gives a percentage, the candidate converts it to a fraction, performs an operation, and forgets that the percentage and the fraction refer to different wholes. A 20% discount applied to a price of 80, followed by adding 20% sales tax to the discounted price, does not return 80. The trap is not in the arithmetic; it is in the assumption that the base of the percentage stayed the same.
Three techniques neutralise the trap. First, anchor every percentage to a concrete whole — write the whole down, then write the percentage as a fraction of that whole. Second, prefer fraction arithmetic to decimal arithmetic when the conversion is clean: 1/3 of 90 is faster to compute as 90 ÷ 3 than as 0.333... × 90. Third, use benchmark fractions (1/2, 1/3, 1/4, 1/5, 1/8, 1/10, 1/12) as a mental anchor when estimation is enough. The GMAT Focus is not a precision contest; it is a calibration contest. Knowing when 0.1667 is close enough to 1/6 saves time that you will need on the harder items.
Percentage questions also surface as word problems about salary increases, price changes, and population growth. A useful rule: when a problem involves two successive percentage changes, replace them with a single equivalent factor. Two successive increases of 10% are equivalent to a single 21% increase, not 20%. Two successive discounts of 15% and 10% are equivalent to a single 23.5% discount. The arithmetic is trivial once you set the problem up as multiplication of factors, and the technique generalises to three or four successive changes that would be tedious to compute step by step.
Ratio, proportion, and mixture problems: the lever-and-pulley family
Ratio questions on the GMAT Focus range from simple 'split 120 in the ratio 3:5' stems to layered mixture problems involving solutions, alloys, or cost blends. The lever-and-pulley metaphor is useful: ratio questions reward you for picking the right total and then allocating the parts. The most common error is letting the answer choices seduce you into checking a total rather than a ratio, or vice versa.
Mixture problems are the highest-altitude form of the family. A 40-litre solution of acid is 20% acid by volume. Some of it is drained and replaced with pure water, ending up at 10% acid. How much was replaced? The trap is to do the percentage arithmetic on the volumes. The cleaner approach is to track the absolute amount of acid. Start with 8 litres of acid, end with 4 litres, so 4 litres of acid were lost in the drained portion. Each litre of drained solution carried 20% acid, so the drained volume is 4 ÷ 0.20 = 20 litres. The whole problem reduces to two percentage translations and a single division.
Weighted averages are a sibling of mixture problems and appear at almost every score band. The alligation method — placing the two component values on the corners of a cross, the overall average in the middle, and reading off the inverse ratio of distances — solves most weighted-average stems in under a minute. Candidates who have not internalised alligation tend to set up simultaneous equations, which works but eats time. For most candidates I work with, the alligation cross becomes the default for any question of the form 'blend X and Y to get Z'.
Rate, time, and work: where word problems get long
Rate problems look like the longest items on the section, partly because the test uses them as scaffolding for multi-step stems. The good news is that the underlying equation is always the same: rate times time equals the quantity produced, covered, or completed. Once that equation is written down, the rest of the problem is usually translation. The bad news is that translation is where candidates lose points, especially when a problem introduces two workers, two machines, or two pipes with different rates.
The cleanest way to attack a work-rate problem is to compute each worker's individual rate first, then add them. If Pipe A fills a tank in 6 hours and Pipe B fills it in 4 hours, A's rate is 1/6 tank per hour and B's is 1/4. Together they fill 1/6 + 1/4 = 5/12 of the tank per hour, and the tank is full in 12/5 hours. The fraction arithmetic is exactly the skill from the conversion family above; work-rate is arithmetic wearing a word-problem costume.
For motion problems, the same D = R × T equation applies, with the extra wrinkle that 'rate' can be a relative rate when two objects move toward or away from each other. Two trains 240 km apart moving toward each other at 60 km/h and 80 km/h meet after 240 ÷ 140 hours. The relative-rate translation is the make-or-break move. Candidates who try to set up separate position equations for each train waste minutes that the adaptive section will not give back.
Pacing and the option-c stalking problem
The single most expensive arithmetic mistake on the GMAT Focus is not a maths error — it is a stalking error. 'Option-c stalking' is the habit of trusting the middle answer choice more than the data warrants. On arithmetic items, the answer choices are usually designed so that the most natural computational path leads to a choice that looks correct but is wrong, while the correct answer is buried in one of the other letters. A candidate who lands on choice C, recognises it as a number they have seen before, and selects it without re-reading the stem is the textbook victim of stalking.
The pacing budget for an arithmetic item depends on where it falls in the section. The first 6–7 questions of Quant carry disproportionate weight because they determine module branching. A reasonable rule of thumb: spend up to 2 minutes 30 seconds on the first six items, including reading the stem carefully and checking the answer against the question's actual ask. On later items, the budget tightens to about 1 minute 45 seconds, and a candidate who is over the budget should make a decision, flag the item, and move. The section punishes lingering far more than it punishes a single missed arithmetic item.
Estimation is the lever that protects pacing. Many arithmetic items can be solved within 5% accuracy, and the answer choices are typically spaced wider than that. A candidate who can estimate 17% of 250 as 40-something, glance at the choices, and pick the one in the 40s saves a full calculation. The candidates I work with who climb from 78 to 83 in Quant almost always shift their time budget toward estimation and away from full-precision calculation. The exception is when the answer choices are very close together — that is a signal to switch into precision mode.
Common pitfalls and how to avoid them
Five pitfalls account for most arithmetic errors in the section. Each one has a defensive habit attached.
- Percentage base confusion: always write the base down before computing. A 20% increase followed by a 20% decrease is not a wash.
- Boundary-case neglect in number properties: when the stem says n is a positive integer, test n = 1. When it says two-digit, test the smallest and largest values. Boundary cases reveal whether the answer holds generally.
- Option-c stalking: if you arrived at the answer through a single calculation, re-read the question. Many arithmetic items have a second read that changes the meaning.
- Unit mismatch in rate problems: km/h and minutes do not mix. Convert before you compute, or use a fraction-based setup that hides the units.
- Premature calculator use: the on-screen calculator is not a substitute for thinking. Reaching for it on 1/3 × 90 costs more time than the division does.
The defensive habits are not free. They add ten to twenty seconds per item in the early weeks of preparation, which can feel like a lot. By week three, the habits disappear into the workflow, and the time cost falls to near zero while the error rate stays low. Most candidates I work with recover the lost seconds within ten days of structured practice.
A 21-day arithmetic repair plan
The plan below is what I would build with a candidate who has already taken a diagnostic and knows that arithmetic is the weakest Quant area. It is dense, but it fits inside a working professional's calendar and assumes roughly 90 minutes of focused study on weekdays and longer sessions on weekends.
| Week | Focus families | Daily drill | Review target |
|---|---|---|---|
| Week 1 | Number properties; fractions, decimals, percentages | 20 mixed arithmetic items, untimed | Log every error by family and root cause |
| Week 2 | Ratio, proportion, mixture, weighted average | 15 items per day, timed at 2 minutes each | Cross-check estimation vs exact computation |
| Week 3 | Rate–time–distance; work-rate; percent change chains | 20 timed items + 5 Data Sufficiency stems | Identify the top two residual error families |
| Week 3, day 7 | Full mixed set, 31 items, section-realistic pacing | Simulate the full Quant section | Score review and re-target |
The plan above is not a substitute for a full GMAT Focus preparation cycle, and it assumes the candidate has already covered Data Interpretation, the two critical-reasoning-adjacent reading items, and the test's interface. Arithmetic is one of three legs of Quant. But it is the leg that, in my experience, generates the highest return on time invested in the first three weeks of preparation.
How arithmetic feeds into your overall Quant score
The Quant section on the GMAT Focus is reported as a single scaled score, with no sub-scores for arithmetic specifically. That means an arithmetic error costs the same scaled point as an algebra error or a geometry error. The corollary is that strengthening arithmetic is a direct path to a higher Quant score, without the noise of balancing many sub-areas. A candidate who has fully internalised the six families above can expect to spend less than 60% of the section's time on arithmetic stems and still get most of them right, freeing time for the more reading-heavy Data Interpretation items and the trickier algebra stems.
The score-band math is also worth internalising. Moving from 78 to 83 typically requires around five additional correct items across the section, depending on which items are missed. Because the section adapts, those five extra correct items come disproportionately from the harder module. To reach the harder module, the first 6–7 items — which are heavy on arithmetic — must be clean. The arithmetic family is therefore the gatekeeper, not just a contributor. Preparation strategies that skip arithmetic in favour of the more exotic-looking items are gambling with the section's branch point.
For candidates aiming at the top score bands, arithmetic preparation is also what makes the difference between solving a hard stem in two minutes and solving it in three. The hard arithmetic items are usually a layering of two families — a rate problem with a percentage twist, a mixture problem with a ratio constraint. The candidate who knows each family cold can read through the layers; the candidate who has to re-derive the family logic under pressure falls behind the clock.
Bringing it together: a working approach for the next prep cycle
Most candidates reading this are probably about to start — or restart — a GMAT Focus preparation cycle. The arithmetic cluster is where I would direct the first three weeks of study for any candidate who scored below 80 on a diagnostic. The first week should be diagnostic and error-logging, the second week should be targeted family drilling with estimation habits in place, and the third week should be timed mixed sets that approximate the section's pacing pressure. By the end of week three, the six families should be recognisable on sight, the conversion traps should be reflexive, and the option-c stalking habit should be broken.
A short checklist of habits worth installing before test day: always anchor a percentage to its base; always test a boundary case on a number-property stem; never trust an answer choice that came from a single quick read; convert units before you compute; prefer fraction arithmetic to decimal arithmetic when the conversion is clean; estimate before reaching for the on-screen calculator. None of these habits is novel, but together they form a working system that protects against the specific errors the GMAT Focus is designed to elicit.
TestPrep İstanbul's arithmetic diagnostic is a natural starting point for candidates who want a structured read on which of the six families above is leaking the most points in their current preparation cycle.