How to triage LSAT Reading Comp questions before you read

LSAT Reading Comprehension is the section most candidates underestimate on the way in and remember on the way out. The four passages look like something you would skim in a graduate seminar, but the questions are not asking what you remember. They are asking what the author means, what would weaken the central inference, and which line performs the structural work of the argument. Preparation strategy for this section is therefore not a vocabulary exercise and not a reading-speed exercise. It is a question-type triage exercise wrapped around a passage map you build in the first 90 seconds.

The LSAT Reading Comprehension format has been stable for decades: roughly 27 questions distributed across four passages, one of which is a comparative reading pair of two shorter texts. You have 35 minutes, which works out to 8 minutes 45 seconds per passage block. The questions sit in three loose families: global, local, and inference. Most candidates lose their score ceiling in the global family, not because the questions are hard, but because they read the passage as if it were a content recall test and then try to retrofit a global claim onto whatever they remember.

The first-sentence contract: what the opening line actually promises you

Every LSAT Reading Comprehension passage opens with a sentence that does one of three things, and recognising which one in the first 30 seconds changes everything that follows. The first sentence either states a thesis, frames a debate, or sets a scope. Most candidates treat all three as the same thing, which is the first mistake that compounds into lost questions later.

A thesis opening tells you what the author is about to argue. The remaining paragraphs will be evidence, concession, and rebuttal. Your job is to identify the load-bearing claim so that any global question that asks for the main point has a target. A debate opening gives you two positions and tells you the author will adjudicate between them. Your map needs to track which side the author takes and why. A scope opening is the trickiest. The author is not arguing a thesis or refereeing a debate. They are defining the boundaries of a phenomenon: what counts, what does not, and what counts only under specific conditions. Global questions on scope passages test whether you noticed the boundaries, not the topic.

Here is the diagnostic move I use with private students. After the first sentence, stop reading and write a one-line prediction: 'This passage is going to argue X.' If the sentence is a thesis, write the thesis. If it is a debate, write 'Author will take Side A over Side B because Y.' If it is a scope, write 'Author is going to define Z in terms of A, B, and not C.' That one-line prediction is the anchor for every global question that follows. Without it, you will be reading the passage as a content dump and then scrambling at question 7 to reconstruct what the author actually thought.

The practical effect on your score: candidates who skip the first-sentence contract average 5 to 7 wrong on the section, and almost all of those errors sit in the global question family. Candidates who build the contract average 3 to 4 wrong, with the residual errors concentrated in inference rather than global. That is the difference between a 165 and a 170 on the LSAT, holding Logical Reasoning and the other scored sections constant.

The four passage archetypes and their question distributions

LSAT Reading Comprehension passages fall into four recognisable archetypes, and each one has a slightly different question distribution. Reading the archetype in the first paragraph lets you budget your time and your attention.

Hard-science passages (law-of-nature, biology, geology, physics) tend to run heavy on local detail questions and relatively light on inference. The author is usually explaining a mechanism, and the questions test whether you understood the mechanism. Expect 6 to 7 local questions, 1 to 2 inference questions, and 1 to 2 global questions. The danger here is retention. The detail is genuinely hard to remember, so the questions reward a strong passage map more than raw reading speed.

Social-science passages (economics, sociology, political science, history) are the most balanced. Expect 3 to 4 global, 3 to 4 local, and 2 to 3 inference questions. The author is usually advancing a claim about a pattern, and the global questions test whether you can name the pattern. This is the archetype where the first-sentence contract pays off most, because the claim is almost always stated up front.

Humanities and law-related passages (legal theory, philosophy, literary criticism) lean heavily on inference. Expect 4 to 5 inference questions, 2 to 3 global, and 2 to 3 local. The author is doing conceptual work: defining a term, drawing a distinction, or arguing that two concepts are not the same. Inference questions on these passages test whether you can extend the author's distinction to a new case, which is a different skill from recognising a pattern.

Comparative reading pairs (two shorter passages on a shared topic) shift the question distribution sharply. Expect 4 to 5 questions that ask you to compare the two authors, 1 to 2 questions on each passage individually, and 1 to 2 on the relationship between them. The pacing budget for a comparative pair is tight: roughly 3 minutes 30 seconds for each short passage, leaving 1 minute 45 seconds for the comparison questions. Candidates who try to read comparative passages at the same depth as the three long passages run out of time on question 22 and start guessing.

Archetype	Typical question mix	Dominant question family	Time budget
Hard science	6–7 local, 1–2 inference, 1–2 global	Local detail	8:45
Social science	3–4 global, 3–4 local, 2–3 inference	Global thesis	8:45
Humanities / law	4–5 inference, 2–3 global, 2–3 local	Inference extension	8:45
Comparative pair	4–5 comparison, 1–2 each passage, 1–2 relationship	Cross-author	3:30 + 3:30 + 1:45

The table is a planning tool, not a rule. Real passages blend archetypes, and a social-science passage can carry 3 inference questions. But the mix tells you where to spend your second-pass time. If you are sitting on a hard-science passage with 7 local questions, do not waste 90 seconds hunting for a thesis in paragraph 4. The thesis was probably stated in paragraph 1 and the rest is mechanism.

The global-reference trap and why it is invisible until question 8

Here is the trap that costs most candidates 3 to 5 questions per section. LSAT global questions do not ask what the passage is about. They ask what the passage is about relative to the question stem. The reference point is the stem, not the passage, and that inversion breaks candidates who trained themselves to read the passage first and then find the answer.

Consider a passage about whether judicial review is a democratic check or an anti-democratic power. The global question is not 'What is this passage about?' It is 'Which one of the following most accurately describes the passage?' The wrong answers are about judicial review. The right answer is the answer that captures the author's specific claim about judicial review, which is a much narrower description. Candidates who skim the passage and pick the answer that mentions judicial review will be wrong, because three of the four answers mention judicial review. Only one describes the author's actual position.

The fix is to read the question stem before you answer, every single time, even for global questions. The stem tells you what kind of claim you are looking for: main point, primary purpose, tone, structure, or relationship to another field. The answer must be a claim of that kind, stated in the author's voice, and the four distractors will be claims of the wrong kind. A main-point question will offer you a primary-purpose distractor. A primary-purpose question will offer you a main-point distractor. Reading the stem first collapses the search space.

This is the part of preparation that candidates skip. They drill passages and skip the taxonomy of global questions, which is a mistake with a clear cost. Of the 27 questions on the section, 8 to 10 are global, and 3 to 5 of those 10 will be global-reference questions where the stem reframes what the passage is about. Without the stem-first habit, you will answer those 3 to 5 questions by gut feel and lose 1 to 2 of them. With the habit, you will answer them by category match and lose 0 to 1.

Question family triage: the 35-minute budget that actually works

Time pressure on LSAT Reading Comprehension is real but negotiable. 35 minutes for 27 questions is 77 seconds per question, but the distribution is not even. Comparative pairs take longer per question because of the cross-author reasoning. Hard-science passages take longer on the first read because of the mechanism. Humanities passages take longer on the questions because of the inference chains. A flat 77-second budget ignores all of this and produces a familiar failure mode: candidates finish three passages comfortably, hit the comparative pair with 9 minutes left, and start guessing at question 20.

The triage I teach starts before the section begins. Rank the four passages in your expected difficulty order based on the first paragraph and the topic. Put the comparative pair second, not fourth, regardless of your preference, because it has the highest variance in question difficulty and benefits from a fresh attention budget. Put the hardest long passage first, when your focus is highest. Put the easiest long passage third, when you are warmed up but not yet fatigued. The fourth slot is the comparative pair's 1 minute 45 seconds of comparison questions, which you should reserve as a buffer rather than spend on the long passages.

Within each passage, the question-by-question budget matters more than the passage-level budget. Local detail questions should take 30 to 45 seconds. Global questions should take 60 to 90 seconds, because they require you to return to the passage map. Inference questions should take 60 to 90 seconds, because they require you to chain two claims. Comparison questions should take 45 to 60 seconds each, because the cross-author claim is usually stated explicitly. If a question has taken 100 seconds and you have not narrowed to two answers, mark it, move on, and return with the remaining 90 seconds in the buffer. The buffer is not optional. It is the difference between a clean finish and a panicked finish.

One number worth holding: a candidate targeting a 170+ LSAT score is missing roughly 3 to 5 questions per Reading Comprehension section, which means 22 to 24 correct out of 27. If you are missing 6 or more on practice tests, the fix is not a faster reading pace. The fix is the question-family triage, because the errors are concentrated in one or two families, not spread evenly. A diagnostic that breaks your misses down by family will tell you within two practice tests whether your ceiling is being set by global, local, or inference errors, and the study plan follows from that.

Common pitfalls and how to avoid them

Five patterns account for most of the avoidable errors I see in private LSAT Reading Comprehension tutoring. None of them are reading-speed problems. They are all mapping and triage problems.

Reading for content instead of structure. Candidates who summarise each paragraph in a content word ('whales', 'court', 'experiment') lose global questions because they cannot recall the author's claim. The fix is to summarise each paragraph in a claim word ('whales communicate through signature calls', 'court rejected the democratic objection', 'experiment suggests X is a cause of Y'). Claim words are retrievable. Content words are not.
Answering the question you wish had been asked. On inference questions, candidates pick the answer that is true in the real world instead of the answer that follows from the passage. The LSAT is not testing whether the inference is true. It is testing whether the inference is supported by the passage. Run every answer back to a specific line. If you cannot point to the line, the answer is wrong, no matter how true it sounds.
Confusing the author's view with a cited view. Humanities and law passages quote other scholars. Candidates who attribute the quoted view to the author lose the global question. Track every cited view with a tag: 'Author says X. Scholar Y says X. Author disagrees.' The tag is what the question is testing.
Skipping the comparative passage relationship map. On comparative pairs, candidates read both passages carefully and then guess on the relationship questions. Build a one-line relationship map during the first read: 'Author 1: claim A. Author 2: claim B. They agree on C, disagree on D.' The relationship questions are then mechanical.
Spending the buffer on a single hard question. If a question has taken 100 seconds and you are still not sure, mark it. The expected value of a third minute on one question is lower than the expected value of a confident answer on the next question. The buffer exists to absorb exactly this trade-off.

How to study Reading Comprehension differently from Logical Reasoning

Logical Reasoning and Reading Comprehension share a question-stem vocabulary, but the study methods diverge in a way most candidates do not notice. Logical Reasoning is argument-by-argument: each question is a self-contained stimulus, and you can drill one argument type per session. Reading Comprehension is passage-shaped: the argument lives inside four paragraphs and 800 words of context, and you cannot drill it one question at a time. The minimum practice unit is a full passage with all of its questions, untimed, with a written passage map after the first read.

The untimed passage map is the single highest-return study activity for this section. Pick a passage you got 4 out of 7 on. Read it fresh, untimed, and write a four-to-six-line passage map: one line per paragraph, plus a final line for the author's overall claim. Then re-attempt the questions you missed. If the map did not change your answers, the errors are content errors, and the fix is more reading. If the map changed your answers, the errors are map errors, and the fix is more mapping. Most candidates discover they are losing questions to map errors, not content errors, which is good news, because mapping is a trainable skill and content is not.

The other study method that pays off is targeted question-family drilling. Take 20 global questions from past sections. Do them in a single sitting, untimed, with the stem-first discipline. Track which global family each question belongs to: main point, primary purpose, tone, structure, or relationship. After 20 questions, you will see a pattern in your misses, and the pattern will tell you which sub-family to drill next. Candidates who do this for two weeks typically gain 2 to 3 questions per section, which on the LSAT scoring scale is the difference between a 168 and a 170.

Integrating Reading Comprehension into a wider LSAT preparation strategy

Reading Comprehension is the third of three scored sections on the LSAT, alongside Logical Reasoning and the unscored variable section, and it is the section that most candidates leave for last in their preparation. That is a mistake. Logical Reasoning has a higher raw question count (roughly 50 to 53 questions across two scored sections) and therefore a higher weight in your scaled score, but Reading Comprehension has a higher ceiling per question. A single Reading Comprehension question is worth the same scaled points as a single Logical Reasoning question, but the difficulty variance is wider, which means a strong Reading Comprehension performance is a tiebreaker at the top of the score band.

The practical implication for your study plan is to interleave Reading Comprehension with Logical Reasoning from week one, not to sequence them. Two passages per week of Reading Comprehension study, plus two Logical Reasoning sections, will build both skills at the rate the LSAT actually tests them. A candidate who does six weeks of pure Logical Reasoning followed by three weeks of Reading Comprehension will have a brittle Reading Comprehension score, because the inference and global-reasoning habits from Logical Reasoning do not transfer automatically into passage context.

One more tactical note on the LSAT exam format. Reading Comprehension is the only section where you can lose points by reading too carefully. Logical Reasoning rewards line-by-line attention. Reading Comprehension rewards passage-map attention, and over-reading a passage costs you time that you cannot recover. The 8 minutes 45 seconds budget is real, and a candidate who reads for 11 minutes on a passage has borrowed 2 minutes 15 seconds from the next passage, which will then be answered under fatigue. The triage discipline is therefore not just a scoring tool. It is a fatigue-management tool, and fatigue is the variable that separates a 170 from a 175 on test day.

Conclusion and next steps

LSAT Reading Comprehension rewards a specific kind of attention: structural, claim-oriented, and stem-aware. The four passage archetypes give you a planning frame, the four question families give you a triage frame, and the global-reference trap is the single largest source of avoidable errors. Build the first-sentence contract, write a claim-word passage map, and read every question stem before you read the passage. Those three habits, practiced over six to eight weeks, will move your section score into the band that holds a 170+ LSAT together.

TestPrep Europe's diagnostic assessment is a natural starting point for candidates who want a question-family breakdown of their Reading Comprehension misses before they build a passage-mapping study plan.

Frequently asked questions

How many LSAT Reading Comprehension questions are global versus local?

On a typical LSAT Reading Comprehension section, 8 to 10 of the 27 questions are global, 10 to 12 are local detail, and 6 to 8 are inference. The exact mix varies by passage archetype, with hard-science passages leaning local and humanities passages leaning inference. Candidates who track their errors by family usually find that one family accounts for 60 to 70 percent of their misses, which makes the study plan straightforward.

Should I read the passage or the question stems first on LSAT Reading Comprehension?

Read the passage first, but only once, and build a claim-word passage map during the read. Then go to the question stems and read each stem before looking at the answer choices. The stem tells you what kind of claim you are looking for, which collapses the four answer choices into two plausible candidates within 10 to 15 seconds. Candidates who skim the passage and dive into the answers typically lose 1 to 2 global questions per section to distractor answers that match the topic but not the stem.

What is the best pacing strategy for the comparative reading pair on the LSAT?

Budget 3 minutes 30 seconds for each short passage and 1 minute 45 seconds for the comparison questions, for a total of about 8 minutes 45 seconds on the pair. Read each short passage with the same claim-word mapping you use on the long passages, and write a one-line relationship map as soon as you finish the second passage. The relationship questions are then mechanical rather than reconstructive, which is the only reliable way to finish the pair with time to spare.

How long should I study LSAT Reading Comprehension before I see score improvement?

Most candidates see measurable improvement after 4 to 6 weeks of structured practice, defined as 2 to 3 full passages per week with written passage maps and targeted question-family drilling. The first sign of improvement is a tighter error pattern: misses concentrated in one or two families rather than scattered across all four. The second sign is a reduction in the number of questions that take more than 90 seconds. Candidates targeting a 170+ LSAT score typically need 8 to 12 weeks of consistent Reading Comprehension work.

Does Reading Comprehension matter more than Logical Reasoning for a high LSAT score?

Logical Reasoning has a higher question count and therefore a higher weight in the raw-to-scaled conversion, but Reading Comprehension has a wider per-question difficulty variance, which makes it a tiebreaker at the top of the score band. A candidate with a strong Logical Reasoning performance and a weak Reading Comprehension performance will plateau around 168 to 170. A candidate with balanced performance across both sections can push into the 172 to 175 band. Treat Reading Comprehension as a ceiling-setter, not a side skill.

How to triage LSAT Reading Comp questions before you read the passage