On this page
The 2026 Reading section at a glance
The TOEFL iBT 2026 Reading section is 30 minutes and 50 items, delivered as two adaptive modules rather than as a fixed set of passages. Module 1 is the same for everyone. How much of it you answer correctly decides whether Module 2 is the harder or the easier form, and only the harder route reaches the top of the band scale, which is why the first half of the section is where your band is really decided.
Three task types appear. Complete the Words gives you a short paragraph with word endings removed and ten blanks to type, one point each, so a single paragraph is worth ten raw points. Read a Daily Life Text is a campus email, notice or message thread carrying two or three questions. The academic passage runs four or five paragraphs and carries five questions. Its passages are drawn from a rotation of academic topics — life sciences, physical sciences, social sciences, arts and humanities, and history — written at the level a first-year undergraduate textbook would use, and they are not adapted from any single source: they are written specifically for the test and exist nowhere else.
The TOEFL iBT reading question types below are the ones that sit behind that academic passage. ETS groups the two factual variants together as key-detail questions, so you will also see these types of reading questions counted as five rather than six. Every item is worth one point, and the section score is reported on the 1.0–6.0 band scale.
1. Factual information questions
Factual information is the most common Reading question type, accounting for roughly a third of the questions on the section. The stem points to a specific paragraph and asks what that paragraph states about a specific topic. The correct answer is a paraphrase of one or two consecutive sentences in the named paragraph.
A direct match for the answer in the named paragraph. If you cannot point to the supporting sentence within thirty seconds, the option you are leaning towards is probably a distractor.
The three classic distractor patterns are paraphrase distortion (real vocabulary from the passage assembled into a claim the passage never made), out-of-scope (a plausible-sounding fact the passage simply does not mention), and partial truth (a claim that would have produced the same outcome via a different mechanism). The full walkthrough at factual-information walkthrough shows one of each on a passage about the history of paper money.
2. Negative factual questions
Negative factual is the same skill as factual, inverted. The stem contains the word NOT or EXCEPT in capitals, and your task is to pick the option that the passage does not state. Three options paraphrase real sentences; one does not. Students who handle factual questions confidently often slip on negative factual because they pick the most familiar-sounding option without checking that it’s the one missing from the passage.
Mark each option as supported (S) or not (N) against the named paragraph. The answer is the one N. Circle the word NOT/EXCEPT in the stem before reading options.
Negative factual questions are typically the question type students lose to time pressure most: they require checking four claims against the passage rather than finding one. Budget 90 seconds for this type rather than the 60 seconds you might budget for a standard factual question.
3. Inference questions
Inference questions ask for something the author clearly implies but never spells out. The correct answer combines two or more pieces of information from the passage to support a small, defensible conclusion. The biggest trap is picking an answer that is stated in the passage — that answer is wrong by definition, because no inference would be needed for it. For a deeper method, the trap patterns, and two worked examples, see the guide to TOEFL Reading inference questions.
An option whose support is two sentences taken together, not a single statement. The correct answer is always the smallest defensible step beyond what the passage states, never a leap.
Eliminate any option that scales up the claim (a specific region becoming the whole brain, the studied group becoming all people) or that flips a stated fact. The inference walkthrough shows all four common inference distractor patterns on a neuroscience passage.
4. Vocabulary in context
Each passage has two or three vocabulary questions. One word or short phrase is highlighted and the question asks which of four options is closest in meaning as the word is used here. The correct answer is rarely the most common dictionary meaning of the word. It is the meaning that makes the sentence still work when you substitute the option for the original.
Substitute each option into the original sentence. The right answer produces a sentence the original author could have written. If a polysemous word has multiple distinct meanings, one wrong option will usually be its other dictionary meaning.
Long-term vocabulary preparation: study the Academic Word List with full ranges of meaning rather than first-meaning-only flashcards. The vocabulary walkthrough uses a Renaissance art passage to show how dictionary thinking gets the wrong answer.
5. Paragraph relationship questions
These ask what a paragraph does in relation to another one, not what it says. The answer is about structure, and the wrong options are usually true statements that describe the wrong structural relationship, which is what makes them hard to reject.
Summarise each paragraph in three or four words describing its job, not its content: "describes the mechanism", "gives an exception to it", "supplies the evidence for it". Then match that pair against the options. A student who summarises what each paragraph is about rather than what it is doing will find two options that both look right.
The six relationships worth recognising are cause and effect, contrast, example, elaboration, qualification, and problem and solution. Our paragraph relationship lesson works each one against a real passage.
6. Important idea and summary questions
These ask which option best expresses the main idea of the passage. In 2026 this is a single-answer multiple choice worth one point, not the old multi-select summary task.
Apply the coverage test. The right answer has to account for the whole passage, not one strong paragraph of it. Wrong options here are almost always true, which is exactly why they are chosen: they are drawn from a single paragraph and presented as though they covered the argument.
Build the thesis in your own words before you read the options, or the first true-sounding option will anchor you. Our important idea lesson covers the coverage test in full.
Two TOEFL question types ETS removed for 2026
If you have prepared from older books or from another platform, you will have practised two question types that the 2026 format no longer contains. Neither appears in the ETS 2026 test specifications, and neither is in the list above.
Insert Text. You were given an extra sentence and four squares placed inside the passage, and you chose which square the sentence belonged in. Picking a square made the sentence appear there so you could read the paragraph with it in place. It tested cohesion: the inserted sentence usually carried a pronoun or connector that needed a referent in the sentence before it.
Prose Summary. The final question of each passage gave you an introductory sentence and six choices, and you selected the three that together summarised the passage. It was the only question worth more than one point, carrying up to two.
Students do occasionally report meeting an unfamiliar question on test day, including something resembling one of these. The likeliest explanation is in ETS’s own blueprint: the reading and listening sections carry non-scored trial items, included so ETS can measure how a question performs before using it for real. Those items do not count towards your score and are not in the published specification, which is why no practice platform contains them. If you meet one, answer it and move on.
Our older walkthroughs for the insert sentence walkthrough and prose summary walkthrough tasks are still online for anyone comparing the two formats, but neither task appears on the 2026 test and neither is worth practice time now.
Time strategy for the whole section
The Reading section gives you 30 minutes for 50 items across two adaptive modules, which is roughly 36 seconds an item. Most students who run out of time did not read slowly; they spent four minutes on one paragraph early on and never recovered it.
Spend your time where the points are. A Complete the Words paragraph carries ten raw points and takes a couple of minutes, which is more than a whole academic passage question set is worth, so it is the last thing to rush. On the academic passage, aim for around 90 seconds on important idea and paragraph relationship questions, and 60 seconds on factual, negative factual and vocabulary. Mark anything that takes more than two minutes and come back if you have the buffer.
If you are short at the end, attempt every remaining letter-completion blank before returning to the academic passage questions. Both are worth one point each, but a blank you can read takes seconds and a passage question does not. Nothing is deducted for a wrong answer, so leaving an item empty guarantees the only score you cannot improve on.
The biggest single time-leak is re-reading whole paragraphs to answer one question. Train against it with the sample walkthroughs above, then take timed Reading tests until the rhythm is automatic.
Try a free Reading practice test
Twenty full Reading practice tests on TOEFLMock, scored on the new 2026 1-6 band scale. First test is free.
Start a Reading practice test