GMAT Verbal Reasoning Practice Test

The format, precisely

Twenty-three questions in 45 minutes, which is 1 minute 57 seconds each and makes this the tightest per-question budget on the exam. Two formats appear. Critical Reasoning gives you a short argument and asks a single question about it, with five choices. Reading Comprehension gives you a passage on the left of a split screen and a question with five choices on the right, then several more questions against the same passage.

The removal of Sentence Correction changed the character of the section completely. It used to be possible to bank a reliable block of marks through memorised grammar, which suited candidates with good recall and rewarded drilling. That route is closed. What remains rewards a repeatable reading procedure, which is slower to build and considerably more durable once built. It also means that verbal preparation for non-native English speakers now looks much more like verbal preparation for native speakers than it used to.

Critical Reasoning

A Critical Reasoning item is an argument of a few sentences followed by a question line. Your job is to do one specific thing to that argument: find what it assumes, find what would damage it, find what would support it, find what would help you evaluate it, or describe the role two marked portions play in it. There are five answer choices and only one is correct.

Measured across 181 official Critical Reasoning items, this is the tightest distribution in the whole corpus. Half of all items run between 79 and 95 words including the question line, with a median of 87 and a full range of 55 to 139. Crucially, the median is 86 at easy, 88 at medium and 87 at hard, so length carries no information about difficulty at all. The reasoning gap is where the difficulty lives.

That tightness is directly usable. A Critical Reasoning read is close to a fixed cost, so you can plan it: around forty to fifty seconds for an 88 word argument read properly, leaving roughly a minute for the choices. If you routinely exceed ninety seconds before touching the choices, you are re-reading, and re-reading is almost always a symptom of having read the question line last rather than first.

The choices are also measurably uniform: across the 341 complete verbal choice sets, the typical item's middle choice runs 16 words, and the gap between its longest and shortest choice is 8. There is no length tell. Only 18 per cent of those sets open with the same word across all five, which tells you something else useful: the choices are parallel in function, not in surface grammar. Do not expect them to look alike, and do not treat a differently phrased choice as suspicious.

The prompt families that recur

A handful of question lines account for most Critical Reasoning items, and they are phrased near-identically each time. Recognising the family before you read the argument is the single highest-return habit in this format, because each family requires a different kind of attention.

• Which of the following, if true, most seriously weakens the argument?
• Which of the following, if true, most strengthens the argument?
• Which of the following is an assumption on which the argument depends?
• Which of the following would be most useful to establish in evaluating the argument?
• In the argument given, the two portions in boldface play which of the following roles?

Weaken and strengthen questions ask you to add a new fact and judge its effect, so the correct choice will usually introduce something the argument never considered. Assumption questions ask what the argument already relies on without stating, so the correct choice contains nothing new; if a choice adds information, it is almost certainly not the assumption. Evaluation questions ask which unknown, once resolved, would most change your confidence, which means the correct choice is a question rather than a fact. Boldface questions ask about structure alone and can often be answered without fully understanding the subject matter.

Read the question line first, then the argument. That order sounds trivial and it changes the read entirely, because you know what you are looking for while you are looking. A weaken question makes you hunt for the gap between evidence and conclusion; an assumption question makes you hunt for the unstated bridge. Reading the argument cold and then discovering what was wanted means reading it twice, and at 1 minute 57 per question you cannot afford that on 23 questions.

Why wrong choices are wrong

The official explanations name a specific error for every wrong choice, and the same handful of mechanisms recur constantly. Learning them by name is faster than deriving each elimination from scratch.

• True but irrelevant. A statement the argument supports which does not answer the question asked. The most common trap for careful readers.
• Too strong. The right idea overstated, usually with always, only, proves or eliminates. Correct answers in this section are rarely absolute.
• Out of scope. Plausible real-world reasoning the argument never licenses. Your outside knowledge is not evidence here.
• Reversed causation. Swapping cause and effect in an otherwise sound reading, which is specific to Critical Reasoning and very effective.
• Right direction, wrong target. Something that weakens or strengthens a different claim from the one the argument actually makes.

Reading Comprehension

A passage sits on the left of a split screen, sometimes with a line-number gutter, and questions appear one at a time on the right. Some questions highlight a portion of the passage in yellow and ask about that portion specifically. You cannot see all the questions for a passage at once, and you cannot return to an earlier one during the section.

The median passage figure hides the most useful fact, which is that passages are not distributed around a middle. They cluster into two distinct lengths: a short form around 210 to 240 words and a long form around 350 to 380. There is very little in between. That means the read is one of two known costs, roughly a minute for a short passage or nearer two for a long one, and you can recognise which you are dealing with the moment it appears.

The exam attaches no more than four questions to a passage, even though the Official Guide prints sets of up to eight for practice purposes. Three or four is the working assumption. That matters for the economics of the read: a two minute investment in a long passage is recovered across four questions, so reading properly the first time is nearly always cheaper than skimming and returning.

Reading Comprehension has the most parallel choices in the corpus. Twenty-three per cent of choice sets share an opening word, and the gap between longest and shortest choice has a median of only 6 words. This is the format where a conspicuously long correct answer would be the loudest possible tell, which is precisely why the official items keep the spread so tight, and why we hold ours inside the same band.

Question stems here are short, a median of 18 words, and they fall into recognisable kinds: main idea, a stated detail, an inference the passage supports without stating, the function of a particular sentence or paragraph, and application of the passage's reasoning to a new case. The distinction that costs the most marks is between a stated idea and an inferred one. A stated-idea question wants something the passage says; an inference question wants something the passage makes necessary. Answering an inference question with a stated fact is a very common and entirely avoidable loss.

What the section actually measures

The official score report tags every verbal question with a question type and a fundamental skill. Question type is simply Critical Reasoning or Reading Comprehension. The fundamental skills are Analysis/Critique, Plan/Construct, Identify Stated Idea and Identify Inferred Idea, and they cut across both formats, which is why the report can tell you something the raw format split cannot.

Underneath, the concept tagging in the official corpus we measured shows where the weight sits. Argument Construction accounts for 76 items and Argument Evaluation for 71, making argument work the dominant activity in the section. Inference follows at 53 and Supporting Idea at 42. Evaluation and Evaluation of a Plan account for 34 each. Main Idea contributes 21 and Application 11.

The practical reading of that table is that inference, in one form or another, is the single most transferable skill on offer. It appears as an explicit concept, it underlies most assumption and strengthen work, and it drives the harder Reading Comprehension questions. A candidate who improves at distinguishing what a text states from what it makes necessary improves across both formats at once, which is not true of most other things you could practise.

Pacing the tightest section on the exam

One minute 57 per question on average, but the work is lumpy. A Reading Comprehension passage takes one to two minutes before you answer anything, and that time is repaid across three or four questions. A Critical Reasoning item is self-contained and should run close to the average. Plan the section as a pool rather than a metronome.

1. Read the question line first on Critical Reasoning, always. It determines how you read the argument.
2. Read a Reading Comprehension passage once, properly, for structure and the author's position rather than for detail. Details can be located later; structure cannot.
3. Predict the answer before looking at the choices where the format allows it. A prediction makes elimination fast and resists attractive wrong answers.
4. Eliminate on a named ground. If you cannot say why a choice is wrong, you have not eliminated it, you have merely disliked it.
5. Check at question 12 with roughly 22 minutes left, and again at question 18 with roughly 10.
6. Guess and bookmark rather than fight. Two questions rescued at the end are worth more than one won in the middle.

The prediction step is the one that separates fast verbal candidates from slow ones. Wrong choices in this section are engineered to be attractive, and reading five attractive options with no position of your own is a reliable way to talk yourself out of a correct answer. Forming even a rough prediction first gives you something to compare against, and turns choice evaluation from a debate into a match.

Reading the verbal score report

You get a section score on the 60 to 90 scale with a percentile, then a breakdown by question type and by fundamental skill, a per-question table with response times, and the time-against-outcome scatter. Verbal has no content domain split, because the section has no content in that sense.

The most informative comparison is Identify Stated Idea against Identify Inferred Idea. A large gap between them is common and highly actionable: it usually means you are answering inference questions with things the passage said, which is a procedural error rather than a comprehension failure and responds quickly to deliberate practice. A gap between Analysis/Critique and Plan/Construct usually points at a specific Critical Reasoning family, most often the evaluation and plan questions, which many candidates have simply never practised as a distinct kind.

In the time scatter, look for long-duration wrong answers on Critical Reasoning specifically. Because Critical Reasoning stems are so uniform in length, an item that took you four minutes was not a longer question; it was a question where you got stuck between two choices and kept re-reading. That is the classic verbal time leak, and the fix is a firmer elimination rule rather than faster reading.

Fitting verbal into a study plan

Verbal improves slowly and steadily rather than in jumps, which makes little and often the right pattern. A 45 minute verbal section test two or three times a week, each followed by a proper review of every question you were not certain about, will move the score more reliably than long weekend sessions. Review the questions you guessed correctly with the same care as the ones you got wrong; the response-time column will identify them for you.

Balance it against the other two sections. The quantitative practice test and the data insights practice test each carry their own 60 to 90 score, and Data Insights in particular rewards the same close-reading skill this section builds, especially in Multi-Source Reasoning. Then measure everything together with a full length practice test every two to three weeks, because verbal endurance across 23 questions at the end of a 135 minute sitting is a different thing from verbal accuracy in a fresh 45 minute block.