GMAT Verbal Reasoning is the 23-question, 45-minute section of the GMAT that most candidates underestimate. The section is built on two question families, Reading Comprehension and Critical Reasoning. Sentence Correction, the grammar question type that older prep books spend chapters on, is no longer part of the GMAT. Each remaining family probes a different reasoning muscle: Reading Comprehension tests how a candidate handles dense argument prose at speed, and Critical Reasoning tests how a candidate attacks the architecture of an argument. Understanding those two families, and how the adaptive section responds to your answers question by question, is the first step in any serious preparation plan.
Most students arrive at the section assuming Verbal is something you can “wing” because they read in English every day. In my experience that assumption is the single most expensive mistake in the entire GMAT prep cycle. Verbal rewards pattern recognition, sentence-level discipline, and a quiet ability to hold a 350-word argument in working memory. A candidate who scores 83 in Quant and 76 in Verbal still loses ground on the total, because the three section scores count equally towards the 205-to-805 total and admissions committees read the section profile as well. The rest of this article breaks down exactly what Verbal contains, how it is scored, how the timing should be planned, and where candidates most often bleed points.
The two question families inside GMAT Verbal Reasoning
Verbal Reasoning on the current GMAT is narrower than it was on the legacy exam. The old Verbal section mixed Reading Comprehension, Critical Reasoning and Sentence Correction; the current section contains only Reading Comprehension and Critical Reasoning. A candidate using a 2018 prep book will read about Sentence Correction and may assume it is still tested in some form. It is not, so hours spent drilling it are hours taken away from the two families that are actually scored.
Reading Comprehension on the GMAT is built around multi-paragraph passages, each linked to several questions. Passages draw on business, social science, and the natural and physical sciences. The questions test five moves: identifying the main idea, locating a specific detail, inferring an unstated conclusion, identifying the function of a highlighted sentence, and asking the candidate to strengthen, weaken, or follow the logic of an argument embedded in the passage. The passages demand the ability to track a thesis through several supporting paragraphs in working memory, which is a trainable skill.
Critical Reasoning on the GMAT has not been watered down. Candidates still see short arguments followed by a single question. The canonical CR question types — strengthen, weaken, assumption, flaw, inference, boldface, and evaluate-the-plan — all appear. The trick is that CR is interleaved with RC rather than placed in a clean block, so a candidate who has drilled CR in isolation needs to retrain the eyes to switch gears every few minutes.
Grammar has not disappeared from your preparation entirely, but its role has changed. There are no questions that ask you to edit a sentence. What remains is grammar as a reading skill: recognising what a modifier attaches to, which noun a pronoun refers to, or how two parallel clauses relate is what lets you read a dense RC sentence or a CR answer choice correctly the first time. Treat grammar as a support skill for RC and CR, not as a question family to drill.
- Reading Comprehension: several passages, each with a few questions, dense business and science prose, around 2 minutes per question on average once the passage has been read.
- Critical Reasoning: short arguments with a single question each, about 1.5–2 minutes per item, seven canonical question types.
The 23-question total is short relative to the legacy 36-question Verbal section, and the shorter section is exactly why every individual question matters more. The implication for preparation is straightforward: a candidate cannot afford to give up on either family, and in particular cannot write off Critical Reasoning in the way some legacy candidates wrote off Sentence Correction.
How GMAT Verbal is scored and why scaled scores move in narrow bands
The GMAT reports a Verbal score on a 60 to 90 scale, the same scale used for Quantitative Reasoning and Data Insights. The three section scores count equally towards the total score, which runs from 205 to 805. The exam no longer includes an Analytical Writing essay, and the old Integrated Reasoning section has been replaced by Data Insights, the third scored section.
On the scoring algorithm, the GMAT is computer-adaptive question by question: after each answer, the engine updates its estimate of the candidate's ability and selects the next item from a calibrated bank, and the final estimate is translated into the 60 to 90 scale. Every question counts. GMAC publishes no rule that the opening questions weigh more than the rest, so the practical lesson is consistent accuracy throughout the section, not a special opening block.
For many candidates aiming at competitive programmes, the high 70s to low 80s is the contest zone. A jump from 78 to 83 is rarely a question of “learning more English.” It usually comes from a small set of specific fixes: steadier pacing across the whole section, less reading-time waste on RC, or repairing the assumption-recognition weakness that drags down CR accuracy.
Admissions committees often read the Verbal score as one signal of how comfortably a candidate will handle case discussion and written work. The exact expectation varies by programme, so check each target programme's published class profile rather than relying on a general rule. What is consistent is that a weak Verbal score is hard to hide when the three sections count equally towards the total.
Reading Comprehension: how to read the passage in 90 seconds without losing the argument
Reading Comprehension is the largest single chunk of Verbal, and the one where most candidates bleed time silently. The trap is over-reading. A candidate who treats each passage like a literature seminar will spend 2.5 minutes on the passage and leave only 30 to 45 seconds per question, which is not enough time to handle the harder inference items. A candidate who skims the passage to “save time” will misread the author’s stance and miss detail questions by two options. The working band sits between those two extremes.
A practical 90-second read looks like this. Spend the first 20 to 25 seconds on the first paragraph, because that paragraph carries the thesis, the author’s stance, and often the specific claim the passage will defend or attack. Read for the nouns and the verbs, not the adjectives. Then move paragraph by paragraph, spending 8 to 12 seconds on each supporting paragraph, asking one question: “What is this paragraph doing for the thesis?” The four-paragraph passage is a thesis paragraph, two supporting paragraphs, and a closing paragraph that often states an implication or raises a counterpoint. The candidate’s job is to leave the passage with four sentences in working memory, one per paragraph, that summarise each block in plain English.
On the question side, the GMAT tests five moves. The main idea question asks the candidate to pick the option that captures the thesis in different vocabulary. The detail question asks the candidate to locate a specific claim, often phrased more abstractly in the answer than in the passage. The inference question asks the candidate to draw a conclusion the author would agree with but did not state outright. The function question asks the candidate to identify what a highlighted sentence is doing within the paragraph. And the argument-based question asks the candidate to strengthen, weaken, or apply the logic of an embedded argument. Each of these moves can be trained separately, which is the only way to bring average time per question down to the 2-minute mark.
Worked example: function of a sentence in a four-paragraph passage
Consider a passage where paragraph 1 introduces a new pricing model, paragraph 2 presents a study supporting it, paragraph 3 raises a competing model, and paragraph 4 concludes that the competing model has a flaw. A function question might highlight the first sentence of paragraph 3 and ask what role it plays. The right answer usually says something like “It introduces an alternative explanation that the passage will later challenge,” not “It summarises the competing view in detail.” The wrong answers tend to over-claim (the passage never offers detail) or to misplace the sentence’s job. A candidate who walked into the passage with a four-sentence summary of each paragraph will see this question in under 30 seconds. A candidate who tried to memorise the prose will spend 90 seconds re-reading the paragraph.
Critical Reasoning: recognising argument architecture under time pressure
Critical Reasoning is the most architecturally predictable part of Verbal. Every short argument has a conclusion, a set of premises, and an unstated assumption. The seven question types — strengthen, weaken, assumption, flaw, inference, boldface, and evaluate-the-plan — all ask the candidate to do something specific with that architecture. The candidate who walks into each CR stem asking “What is the conclusion? What are the premises? What is the unstated assumption?” has already done most of the work before reading the options.
Strengthen questions ask for an option that makes the conclusion more likely. Weaken questions ask for an option that makes the conclusion less likely. Assumption questions ask for an option that must be true for the conclusion to follow. Flaw questions ask the candidate to name the gap between the premises and the conclusion. Inference questions ask for something the argument guarantees, not something the argument merely suggests. Boldface questions describe what each of two highlighted statements is doing in the argument. Evaluate-the-plan questions ask the candidate to identify what kind of evidence would settle the question.
The hardest CR items are usually the assumption items, because the right answer is often a statement that the candidate feels is “obvious” and therefore dismisses. A good rule of thumb: the more obvious an assumption feels, the more likely the test is to test it. Weaken items are the second-hardest, because the wrong answers are often plausible reasons to doubt the conclusion, but the right answer is usually a reason that the author would also accept as relevant. Candidates who over-think weaken items tend to import outside knowledge, which the GMAT explicitly does not reward.
Worked example: a weaken stem in 80 seconds
Take a short argument: “Company X's new product line increased revenue by 18% last quarter, so the company's decision to discontinue its older mid-range product was correct.” The conclusion is that the discontinuation decision was correct; the premise is the 18% revenue jump. The unstated assumption is that the revenue jump was caused by the discontinuation rather than by other factors. A right weaken answer would offer an alternative cause for the jump (a competitor exited the market, the new product line itself would have lifted revenue even without the discontinuation, the company raised prices during the quarter). A wrong weaken answer would attack the 18% number without linking the attack to the discontinuation decision. The candidate who has trained to look for the causal link between premise and conclusion will pick the right answer quickly; the candidate who has trained to argue with the premises will spend two minutes and still choose a distractor.
Grammar as a support skill: reading dense sentences correctly the first time
The current GMAT has no grammar questions. But GMAT passages and answer choices are written in dense, carefully qualified English, and a misread sentence costs you the question just as surely as a misread argument. That is why grammar still deserves some attention: not as rules to apply in an editing task, but as a way of parsing what a sentence actually claims.
Three structures cause most misreadings: modifiers (which noun or verb a phrase attaches to), pronoun reference (which earlier noun “it” or “they” points to), and parallel structure (which elements a list or comparison actually joins). A candidate who can spot each of those structures in a quick scan of a sentence will read a long RC sentence or a qualified CR answer choice correctly on the first pass, instead of re-reading it under time pressure.
Modifier attachment is the most valuable of the three to train, because it changes meaning. In Critical Reasoning especially, an answer choice that looks like it weakens the argument can turn out to describe a different event or a different group once you see what the modifier attaches to. A candidate who reads for ambiguity will pick the right answer faster than a candidate who reads quickly and assumes.
Worked example: a misread modifier in a Critical Reasoning answer choice
Take an answer choice such as: “The CEO announced the price cut after publishing the quarterly report to employees in a company-wide email.” Does the email carry the announcement or the report, and were employees the audience for the announcement or only for the report? If the argument is about how customers reacted to the price cut, the answer to that question decides whether the choice is relevant at all. A candidate who scans for what each phrase attaches to will see the problem in under 15 seconds. A candidate who reads past it may pick a choice that sounds relevant but describes a different event.
Time management: building a minute budget that survives the 45-minute clock
The Verbal section gives the candidate 45 minutes for 23 questions, which works out to just under two minutes per question on average. The distribution, however, is not even. RC passages need about 2.5 minutes for the first read, and then about 1.5 to 2 minutes per question, so the RC share of the section can quietly absorb most of the clock. CR items are usually 1.5 to 2 minutes each. Every question counts, so aim for steady accuracy from the first item to the last rather than a fast start or a rushed finish.
A practical minute budget looks like this. Plan to spend no more than 2.5 minutes reading a passage, with the rest of the RC time distributed across its questions. Plan to spend no more than 2 minutes per CR item. Reserve the last 3 to 4 minutes for a calm review of bookmarked items, where you can change up to three answers, not for new attempts. A candidate who runs out of time at item 21 has mis-spent time in the middle of the section; a candidate who finishes with several minutes to spare has probably rushed somewhere.
The single most common pacing error is letting RC passages eat the middle of the section. An RC passage can quietly absorb several extra minutes if the candidate re-reads supporting paragraphs, and by the time the candidate reaches the last Critical Reasoning items there is too little time left. Those items get rushed, and points are lost on questions the candidate could have answered. A pre-built 90-second-per-passage read eliminates this risk.
Common pitfalls and how to avoid them on GMAT Verbal
The most expensive mistake in Verbal preparation is treating the section as a reading-speed exercise. A candidate who reads more in English does not automatically score higher on the GMAT, because the section rewards reasoning discipline, not raw reading speed. The fix is to train each question family separately, with timed drills that isolate one question type at a time, and to combine the two families in the later weeks of preparation.
The second most expensive mistake is using prep materials written for the legacy GMAT. The legacy exam had a 36-question Verbal section that included Sentence Correction, a separate Integrated Reasoning section, and an Analytical Writing essay. The current GMAT has a 23-question Verbal section with Reading Comprehension and Critical Reasoning only, and no essay. A candidate who has drilled 200 Sentence Correction items from a 2018 prep book has practised a question type that is no longer scored.
The third most expensive mistake is rushing the questions that look easy. The GMAT adapts question by question, so a careless miss anywhere in the section steers you towards easier questions, and you then need a run of correct answers on harder items to climb back. A calm, careful pace on every question is more important than a fast start.
The fourth most expensive mistake is letting a bad question bleed into the next one. A candidate who spends 3 minutes on a single RC inference item, gets frustrated, and then rushes the next two items will lose more points on the rushed items than were ever at stake on the original. The discipline is to pick the best answer you can defend, bookmark it, move on, and come back at the end of the section if time allows.
The fifth most expensive mistake is using outside knowledge in Critical Reasoning. The GMAT does not reward domain knowledge. A right weaken answer must weaken the argument using the same premises the author used, not using facts the candidate brought from elsewhere. A candidate who argues with the author’s premises will choose distractors that the test never intended to be right answers.
A focused 8-week preparation plan for GMAT Verbal
An 8-week Verbal plan should be split into three phases: foundation (weeks 1 to 3), integration (weeks 4 to 6), and simulation (weeks 7 to 8). The foundation phase builds the underlying skills. The integration phase combines the question families. The simulation phase rehearses the full 45-minute section under timed conditions.
Weeks 1 to 3 should focus on one skill block per week. Week 1: Reading Comprehension, with daily timed drills on a single passage and three to four questions, plus an untimed review that builds the four-sentence summary habit. Week 2: Critical Reasoning, with daily drills on one question type per day, plus a weekly mixed-CR drill. Week 3: dense-sentence reading, with daily drills that train you to parse modifiers, pronoun reference and parallel structure in real RC sentences and CR answer choices, alongside mixed RC and CR sets. The candidate should aim to log 8 to 10 focused hours per week, with the morning reserved for new drills and the evening reserved for review.
Weeks 4 to 6 should integrate the two families. Each session should mix RC passages and CR items in random order, with the full 45-minute, 23-question clock running once per week. The candidate should keep an error log that classifies every missed item by question type, by error reason (misread, mis-reasoned, ran out of time, second-guessed), and by family. The error log is more valuable than the raw score, because it tells the candidate which family to drill next.
Weeks 7 to 8 should be full-section simulations. Two timed full sections per week, with a 24-hour cooldown between attempts, plus a daily review of one or two missed items. The candidate should sit the simulation in the same seat, at the same time of day, and with the same breaks as the real exam. A simulation taken in a café on a Saturday morning does not transfer to a Monday-morning test centre.
The plan should also include a daily 20-minute reading habit that does not use GMAT-style material. The New Yorker, the Economist, or a long-form business article trains the eyes to track thesis-and-support structure without the pressure of a timer. A candidate who reads 20 minutes a day for eight weeks arrives at the test with stronger working-memory capacity for RC passages, which compounds with the drill work.
How Verbal fits into the rest of the GMAT preparation plan
Verbal is one of three scored sections, alongside Quantitative Reasoning and Data Insights, and the three count equally towards the total. Holding the other two constant, every additional Verbal point raises the total, and the same is true for the other sections.
For candidates balancing Verbal with Quant preparation, the rule of thumb is to allocate study time in proportion to the gap between the current score and the target score. A candidate at 76 in Verbal and 83 in Quant should spend roughly twice as many hours on Verbal as on Quant, because the Verbal gap is twice as large. A candidate at 80 in both sections should split the hours evenly. The principle is that preparation time is a finite resource, and the section with the largest gap to the target is the one that returns the highest score per hour.
You choose the order of the three sections when you sit the GMAT, and most candidates choose the order in which they want to spend mental energy. In my experience, candidates who read best when fresh often place Verbal first, because it demands sustained reading attention. Candidates who warm up slowly often place Quant first. There is no universally right answer, and the choice should be made after two or three full simulations.
Conclusion and next steps
GMAT Verbal is a 23-question, 45-minute reasoning test built on two families, Reading Comprehension and Critical Reasoning, and it rewards a candidate who treats the section as architecture, not as English trivia. The path from 76 to 83 is rarely a question of more reading; it is a question of tighter pacing, better argument modelling, and consistent accuracy across the whole section. An 8-week plan that builds one skill block at a time, integrates the two families in the middle weeks, and simulates the full section in the last two weeks is a structure that moves many candidates into a higher band.
TestPrep Istanbul's diagnostic assessment is a natural starting point for candidates building a sharper Verbal preparation plan.
| Question family | What you see | Target time per item | Highest-value skill to train |
|---|---|---|---|
| Reading Comprehension | Several passages, a few questions each | ~2 minutes per question | Four-sentence passage summary |
| Critical Reasoning | Short arguments, one question each | 1.5–2 minutes per item | Conclusion / premise / assumption mapping |
| Combined section | 23 questions, 45 minutes | Under 2 minutes per item on average | Consistent accuracy on every question |