When to use one rate, when to use the reciprocal

GMAT Focus rate and work problems sit in the Problem Solving portion of the Quant section, where roughly half of the items are pure math stems and the other half are word problems built around real-life scenarios. Rate questions ask how fast something happens; work questions ask how long two machines, two workers, or two pipes need to finish a job together. The two problem types share a single engine, the relationship rate × time = work, but on the exam they are framed with deliberately different distractors. Candidates who treat every rate stem as a single equation race through the easy half of the bank and lose the medium-difficulty items where the test writer has hidden a unit conversion, a fill phase, or a partial completion.

The goal of this article is to give you a setup-first method: a small set of equation families, a habit of writing units next to every number, and a triage rule for the moment you see a stem that mixes two rates. Working through five worked patterns below will give you the muscle memory to recognise the family within 20 seconds of reading the stem, which is the threshold that separates a Quant score in the low 60s from one that breaks 80.

The single engine: rate × time = work, and why units decide your answer

Every rate and work problem on the GMAT Focus reduces to the equation rate × time = work, where work is usually 1 (one job, one pool, one dataset). The trap is not the equation itself; it is the unit of the rate. A pipe that fills at 6 litres per minute and a pool that holds 240 litres are speaking the same language, but the moment a stem mixes hours and minutes, or jobs and sub-tasks, candidates who skip the unit line in their notebook will write a setup that looks reasonable and produces an answer that is off by a factor of 60.

Build the habit of writing the unit next to every number the moment it appears in your notes. If a worker paints 30 square metres per 4 hours, the rate is not 30; it is 30/4 square metres per hour, or 7.5 m²/h. If a tap fills a tank in 50 minutes and another empties it in 80 minutes, the combined rate is 1/50 − 1/80 tanks per minute, not 50 − 80. The minus sign on the second term is the only thing protecting you from the answer choice that quietly assumes both taps help.

A second habit is to convert the question to a single target before you set up the equation. If the stem asks how long until the pool is full, the work is 1 pool. If it asks how much is filled in 30 minutes, the work is whatever the combined rate yields when multiplied by half an hour. The most expensive error candidates make is to solve for the wrong target, then pick a number that fits the equation but not the question. Train yourself to circle the actual question word — how long, how much, how many — before you start writing the equation.

Finally, on the GMAT Focus, every answer is a positive number or a fraction in lowest terms. If your setup yields a negative rate, you have a sign error, not a trick answer. If it yields a rate greater than the larger of the two individual rates in a combined-work question, the most likely cause is that you added rates when you should have subtracted, or you doubled a term that should have stayed single.

Family 1: the lone rate, with a unit conversion in disguise

The first family is a single rate applied to a single job, but with a unit mismatch between the rate and the target. A typical stem gives a machine producing 450 items in 6 hours and asks how many it produces in 40 minutes. Many candidates see 450, 6, and 40 and start dividing. The clean setup is to write the rate as 450 items per 6 hours, simplify to 75 items per hour, and then convert 40 minutes to 2/3 of an hour before multiplying: 75 × 2/3 = 50 items.

Notice what just happened. The problem was not about algebra; it was about reading units. The GMAT Focus rewards candidates who pause long enough to convert minutes to hours, or feet to metres, or dollars per kilogram to cents per gram, before plugging into the rate equation. The distractor answer choices for this family are almost always the result of skipping the conversion: 75 × 40 = 3,000 items (using 40 as if it were an hour count), or 450/40 = 11.25 (treating 40 as the divisor).

For most candidates I work with, this family is the easiest to fix because the fix is mechanical. Take one full timed set of 20 rate items and, for every stem, write the units next to the rate before computing. After two sessions, the habit is in place. The payoff shows up not only on lone-rate questions but on every later family, because every multi-rate problem on the GMAT Focus is a stack of lone-rate problems with a unit bridge between them.

Family 2: combined work with a shared target

The second family is the classic two-worker or two-pipe problem. Worker A finishes the job alone in a hours, worker B in b hours, working together they finish in t hours, and the equation 1/a + 1/b = 1/t sits underneath. The numerical values on the GMAT Focus usually give small denominators (6, 10, 12, 15) so that the combined rate reduces to a clean fraction, but the stem will hide a real-life detail to make sure you are reading the scenario and not just the numbers.

Worked example: Machine X prints 240 brochures in 6 hours; Machine Y prints 240 brochures in 4 hours. How long does the combined run take, assuming each machine works at a constant rate? X's rate is 40 brochures per hour, Y's rate is 60 brochures per hour, the combined rate is 100 brochures per hour, and the time to print 240 brochures is 240/100 = 2.4 hours, or 2 hours 24 minutes. The trap answer is to add the times (6 + 4 = 10) or to take the average of the two times (5), both of which ignore the multiplicative nature of combined work.

When the two rates operate in opposite directions — one fills, the other empties — the equation is the same except for the sign. The combined rate is 1/a − 1/b (assuming a is the slower, the one that needs more time alone, so its rate is the smaller fraction). The result will be a smaller combined rate than the larger of the two individual rates, and if you set up the equation correctly, you will see that the tank never fills at all if the emptying rate is the larger one. Several GMAT Focus stems test exactly this: a partially filled tank, a tap and a drain running together, and the question asks whether the tank eventually fills or empties. The answer is structural, not numerical.

Family 3: the partial-completion handoff

The third family is the one that costs candidates the most points. One worker starts the job, works for a fixed period, then hands off to a second worker. The stem asks for the total time, the remaining work, or the speed of the second worker. The setup is the same engine, but the work term is no longer 1 — it is a fraction of 1 that depends on how much the first worker completed.

Worked example: Worker P can paint a fence in 8 hours; Worker Q can paint the same fence in 12 hours. P works alone for 3 hours and then stops. How many more hours does Q need to finish? P's rate is 1/8 fence per hour, so in 3 hours P completes 3/8 of the fence. The remaining work is 5/8. Q's rate is 1/12 fence per hour, so the time required is (5/8) ÷ (1/12) = (5/8) × 12 = 60/8 = 7.5 hours. The trap answer is 5 hours, which is the time Q would need to paint the whole fence from scratch, not the remaining 5/8.

A second version of this family has the second worker joining partway through, both working together for a while, and then one of them leaving. The same setup applies, but you solve for the unknown by writing the work as the sum of three phases: solo work by worker A, joint work by A and B, and solo work by B. Each phase has its own rate and time, and the sum of the work done across the three phases must equal 1. Candidates who try to compress three phases into one equation usually lose a term; writing them out as three short lines is faster and more reliable.

Family 4: the rate that depends on a second variable

The fourth family introduces a second variable, usually the number of workers, machines, or pipes, and the rate scales linearly with that variable. The equation is now n × r × t = 1, where n is the headcount, r is the per-unit rate, and t is the time. A typical stem says that 5 machines produce 1,000 units in 4 hours, and asks how many units 8 machines produce in 6 hours, assuming each machine runs at the same constant rate.

Setup: 5 machines × r × 4 hours = 1,000 units, so the per-machine rate is r = 50 units per hour. With 8 machines for 6 hours, output is 8 × 50 × 6 = 2,400 units. The distractor here is the proportional reasoning shortcut 1,000 × (8/5) × (6/4) = 1,000 × 1.6 × 1.5 = 2,400, which is mathematically identical but easier to misread when one of the ratios is inverted. If you use the shortcut, write both ratios on the page, not in your head, and label them machine ratio and time ratio so that the inversion error is visible before you pick an answer.

A more difficult version of this family varies the per-unit rate with a condition. For instance, a factory produces 2,400 units when 8 workers each work 6 hours. If the factory adds 4 more workers but each worker can only work 4 hours, what is the new output? The setup is the same: total worker-hours is the conserved quantity at constant per-worker rate. The first scenario is 8 × 6 = 48 worker-hours, producing 2,400 units, so the rate is 50 units per worker-hour. The second scenario is 12 × 4 = 48 worker-hours, producing the same 2,400 units. The answer is not 3,200 units; the constraint is binding. Several GMAT Focus stems hide a binding constraint of this kind, and the only way to spot it is to compute the worker-hours and compare.

Family 5: the conversion rate between two pipelines

The fifth family is the conversion rate between two units of work: a bottling line fills bottles at one rate and a capping line caps bottles at another rate. The system as a whole produces finished units at the rate of the slower of the two stages, but the question usually asks about the time for the slower stage to clear a backlog created when the faster stage ran ahead.

Worked example: A filling line fills 300 bottles per hour; a capping line caps 240 bottles per hour. Both run for 2 hours. How long will the capping line take, running alone, to clear the un-capped backlog? The filling line produces 600 bottles, the capping line caps 480, the backlog is 120 un-capped bottles, and the capping line clears them in 120/240 = 0.5 hours, or 30 minutes. The trap answer is 2 hours, the time both lines ran together, which ignores the rate differential between the two stages.

The broader lesson is that on the GMAT Focus, a stem that names two distinct production stages is testing whether you can identify the binding rate. In any pipeline of k stages, the throughput is the minimum of the k stage rates, and the queue at any non-binding stage grows at the difference between its own rate and the binding rate. Candidates who pick the wrong answer in this family almost always picked the binding rate as the throughput, then forgot to account for the queue that built up at the faster stage before the question's clock started.

Setting up the equation on the first read: a 20-second triage

The reason a small set of families covers almost every rate and work stem on the GMAT Focus is that the test writer's job is to vary the surface, not the engine. A 20-second triage on the first read will classify nearly every stem into one of the five families above, and the classification tells you the setup before you have read the last sentence of the stem.

The triage has four checks. (1) How many actors are doing work — one, two, or a scalable group? (2) Is the question about a single moment in time or about a duration? (3) Are the rates in the same units, or do you need to convert? (4) Is there a partial-completion phase, a handoff, or a backlog from a prior phase? The first two checks classify the stem into a family; the third check protects you from a unit error; the fourth check protects you from the most common trap in the medium-difficulty items.

Common pitfalls and how to avoid them:

Averaging rates instead of summing them. If A does the job in 6 hours and B in 10 hours, the combined time is not 8 hours. The combined rate is 1/6 + 1/10 = 8/30 jobs per hour, and the combined time is 30/8 = 3.75 hours. The averaging error is the most frequent single mistake on combined-work items.
Forgetting the partial-completion phase. When the stem says A worked for 3 hours and then stopped, the work done before the stop must be subtracted from 1 before you divide by B's rate. Skipping this step gives the time B would need to do the whole job, not the remaining fraction.
Mixing units silently. A rate of 30 metres per minute and a target time of 2 hours will produce the wrong answer if you multiply without converting one of them. Write the unit next to the rate and next to the answer, every time.
Adding rates when one is subtractive. A fill rate and a drain rate combined are r_fill − r_drain, not r_fill + r_drain. If the drain is larger, the tank empties, and the answer to a "how long until full" question is "never."
Inverting a ratio in a proportional-reasoning shortcut. When using the shortcut, label each ratio on the page. An inverted ratio produces an answer that is off by a square factor and is hard to detect after the fact.

Calibrating difficulty: how GMAT Focus rates these items in the adaptive module

The Quant section of the GMAT Focus is delivered in a two-stage adaptive format: the difficulty of the second module is calibrated to your performance on the first. Rate and work problems appear at every difficulty band, but the surface complexity of the stem scales with the band. In the easier half of the first module, the items are usually Family 1 (lone rate with a unit conversion) or Family 2 (two workers, shared target, no handoff). In the harder half of the first module and across the second module, the items migrate to Family 3 (partial-completion handoff), Family 4 (rate that scales with a headcount), and Family 5 (two-stage pipeline with a backlog).

What this means in practice is that a candidate who has drilled Families 1 and 2 but never seen Family 3 will answer the first six items quickly and accurately, then stall on items 7 through 10. The adaptive algorithm reads that stall and lowers the second-module difficulty, capping the Quant score below 70. The fix is to spend at least 30% of your rate and work prep on Family 3 items, even though they feel slower and clumsier, because they are the items that determine which adaptive branch you end up in.

Family	Surface signal in the stem	Setup pattern	Watch out for
Lone rate with conversion	One actor, one rate, mixed units	Simplify rate, convert target, multiply	Skipping the unit conversion
Combined work, shared target	Two actors, "together" or "alternating"	1/a + 1/b = 1/t	Averaging times instead of summing rates
Partial-completion handoff	One actor works, then stops or is joined	Sum of two or three phases = 1	Forgetting the phase before the handoff
Scalable group rate	Headcount, machines, or workers scale	n × r × t = work	Inverted ratio or binding constraint
Two-stage pipeline	Two production stages at different rates	Throughput = min(rates); backlog = difference	Forgetting the queue at the faster stage

A 30-question prep block built around the five families

The most efficient way to internalise the five families is to drill them in isolation for a week, then mix them in timed sets the following week. A clean prep block is 30 questions, divided as follows: 6 from Family 1, 6 from Family 2, 8 from Family 3, 6 from Family 4, and 4 from Family 5. The skewed allocation toward Family 3 is intentional; it is the family that decides which adaptive branch you end up in, and most candidates under-train it relative to its weight in the score.

For each question, time the first read: 20 seconds to classify the stem into a family, 30 seconds to write the setup, and 60 to 90 seconds to compute. Total budget is around 2 minutes per item, which is the average pacing on the Quant section once you account for the easier items you will finish in 60 seconds. If you go past 2:30 on a Family 1 or Family 2 item, the diagnosis is usually a unit error or a missing phase; if you go past 2:30 on a Family 3 or Family 4 item, the diagnosis is usually a setup mistake, and the right move is to abandon the algebraic approach and switch to a proportional-reasoning shortcut if the numbers allow it.

For the mixed-timed week, take three official-length Problem Solving sets of about 15 items each, and mark every rate or work item with the family number as soon as you finish it. After each set, count the misses by family. If more than half your misses are in one family, that is where the next week of prep should go. If the misses are spread evenly, the diagnosis is not family-specific; it is pacing, and the fix is a slower first read with the 20-second triage built in.

How scoring and preparation strategy interact for rate items

The GMAT Focus Quant section is scored on a 60-to-90 scale, and within that scale, rate and work problems are not weighted differently from any other Problem Solving item. The reason they deserve a dedicated prep block is that they sit at the intersection of algebraic setup, unit discipline, and word-problem reading. A candidate who is strong in algebra and weak in unit discipline will lose more points on rate items than on pure algebra items; a candidate who is strong in unit discipline and weak in setup will lose more points on rate items than on any other word-problem family.

For most candidates I work with, the right preparation strategy is to delay rate and work drilling until after the algebra base is solid, then spend 8 to 10 hours across two weeks on the five families above. The first four hours should be untimed and family-by-family, with worked examples read twice each. The next four hours should be timed and family-mixed, with the family label written next to every item in the review. The final two hours should be one full-length Quant section under timed conditions, with rate and work items flagged for a second review if any were missed.

One tactical note on scoring: the GMAT Focus does not penalise wrong answers, so the expected-value calculation on a hard Family 3 item is the same as on any other item — guess if you have eliminated one or two choices, skip only if all five choices look equally plausible. The reason this matters for rate items specifically is that Family 3 and Family 4 items are the ones where candidates most often spend 3 minutes and end up with nothing. Building a 2-minute cap into your pacing, and falling back to a guess once you pass it, recovers a measurable number of points over a full section.

Pulling it together: a 90-second worked example end to end

To see the method in action, walk through a single stem from classification to answer. A factory has two machines, A and B. Machine A can produce 1,000 units in 5 hours; Machine B can produce 1,500 units in 6 hours. Machine A runs alone for 2 hours and then stops. Machine B then runs alone. How many hours does Machine B need to finish the remaining units?

Classification check: two actors, handoff after A's solo phase, ask is about B's solo time. This is Family 3, partial-completion handoff. A's rate is 1,000/5 = 200 units per hour. In 2 hours, A produces 400 units, leaving 1,000 − 400 = 600 units for B. B's rate is 1,500/6 = 250 units per hour. Time required is 600/250 = 2.4 hours, or 2 hours 24 minutes. The trap answer is 3.75 hours, which is the time B would need to produce all 1,000 units from scratch. The distractor is built to catch candidates who skip the partial-completion phase.

The total elapsed time on this stem was about 75 seconds once the family was identified, which is well inside the 2-minute budget. The mechanical work was minimal; the value was in the 20-second triage that told the candidate, before computing anything, that the work was 1,000 units, the first phase was A alone for 2 hours, and the second phase was B alone. Without that triage, the candidate would have spent 30 seconds reading, 30 seconds setting up a combined-rate equation that does not apply, 60 seconds computing, and another 30 seconds arguing with the answer choices. The triage is the whole game.

Conclusion and next steps

GMAT Focus rate and work problems reward a setup-first approach: a small set of five equation families, a habit of writing units next to every number, and a 20-second triage that classifies the stem before you compute. The medium-difficulty items — Family 3 handoffs and Family 4 scalable rates — are the ones that determine your adaptive branch, and they are the ones most candidates under-train. A two-week prep block built around the five families, with a deliberate 30% allocation to the handoff family, is enough to move your Quant ceiling by several scaled points.

TestPrep Europe's rate and work diagnostic set is a natural starting point for candidates who want to see which family is currently costing them the most points before they commit to a prep block.

Frequently asked questions

How many rate and work problems appear on the GMAT Focus Quant section?

The GMAT Focus does not publish a fixed count for any sub-topic, but rate and work items typically make up 15 to 25% of the Problem Solving bank across a full prep cycle. They appear in both adaptive modules, with the harder families concentrated in module 2.

What is the fastest way to classify a rate and work stem on the GMAT Focus?

Use a 20-second triage: count the actors, check whether the question is a moment or a duration, confirm the rate units, and look for a handoff or a partial-completion phase. The first two checks place the stem in one of five families; the second two protect you from the most common unit and setup errors.

Should I always use the combined-rate equation 1/a + 1/b = 1/t for two-worker problems?

Only when both workers are doing the same kind of work on the same job for the entire duration. If one worker starts and stops, or if one fills and the other empties, the equation is the same in form but the work is not 1; it is the remaining fraction after the first phase, and the sign on the second rate may be negative.

How much GMAT Focus prep time should I spend on rate and work problems?

For candidates targeting 80 or higher on Quant, 8 to 10 hours across two weeks is a reasonable allocation once the algebra base is solid. Spend the first half of that block untimed and family-by-family, the second half timed and mixed, and finish with one full-length section where rate items are flagged for second-pass review.

What is the most common mistake on combined-work items?

Averaging the two times instead of summing the two rates. If A finishes the job in 6 hours and B in 10 hours, the combined time is 30/8 = 3.75 hours, not 8 hours. The error is structural and accounts for a large share of misses on Family 2 items across most prep cycles.