Must Be True and Inference questions sit at the centre of LSAT Logical Reasoning, and yet most candidates treat them as the soft middle of a section. The stimulus feels conversational, the stem looks like a small request for a summary, and the answer choices read like slight rewrites of the same sentence. The work, in practice, is harder than that surface suggests. These items punish the reader who adds a word, who softens a quantifier, or who treats a defensible interpretation as if it were a guaranteed one. A clean LSAT Must Be True routine is a small set of habits applied with discipline, and the rest of this article walks through what that routine actually looks like at the whiteboard.
What LSAT Must Be True and Inference questions are really asking
The most common mistake I see in the first month of LSAT preparation is treating a Must Be True stem as a reading comprehension question. Candidates read the stimulus, form a personal opinion, then look for the choice that matches that opinion. The problem is that LSAT Must Be True items do not ask what the author probably believes, what follows if we add a charitable reading, or what a smart reader would infer in a normal conversation. They ask one narrow thing: which answer is forced by the stimulus as written, with no extra information, no softening, and no imported word.
On the LSAT, Must Be True questions are usually flagged by stems like 'which one of the following must be true', 'which one of the following can be logically drawn', or 'the conclusion above logically commits the author to which of the following'. Inference questions look almost identical, with stems like 'which one of the following can be inferred' or 'which one of the following follows logically from the statements above'. The two categories are scored identically and the test-makers mix them, but the labels do carry a small signal. 'Must be true' is the stricter demand: the answer is the only choice that cannot be contradicted. 'Can be inferred' is slightly softer, but in practice it still means 'must be true on this test'.
For most candidates reading this for the first time, the practical takeaway is that the difference between a 165 and a 170 on the LSAT often lives in the precision with which these stems are read. If you learn to treat the stem as a contract, you start to triage choices quickly. A choice that adds a word the stimulus never used, that turns a 'some' into a 'most', or that introduces a new comparison is not just slightly off. It is wrong. The LSAT Logical Reasoning section rewards the reader who notices the missing word before noticing the right idea.
The four internal moves of a Must Be True question
When a candidate asks me how to read a Must Be True stimulus, I usually walk them through four moves. The first is to identify the conclusion. Must Be True items often look like arguments, sometimes look like sets of facts, and sometimes look like disputes. In every shape, the conclusion is the sentence the stimulus is trying to justify, and the premises are the sentences doing the justifying. Even when the question asks only what must be true, the conclusion acts as a load-bearing wall. The choices that contradict the conclusion are eliminated first.
The second move is to map the quantifiers. The LSAT uses a small set of quantifier words, and the difference between them decides entire questions. 'All' forces a universal claim. 'Most' allows exceptions. 'Some' guarantees at least one case but says nothing about the rest. 'Many', 'a few', and 'several' are softer still. Candidates who treat 'most' and 'all' as synonyms lose Must Be True points in clusters. A useful drill is to underline every quantifier in the stimulus and then check whether each answer choice preserves the same quantifier strength. The choice that quietly upgrades 'some' to 'most' is a trap answer, and the LSAT writes that trap on purpose.
The third move is to list the entities and the relationships. In my experience, drawing a tiny diagram on the scrap paper, even a single line connecting two abstract labels, prevents the reader from smuggling in a relationship that the stimulus never stated. If the stimulus says 'critics praised the novel', the only relationship you can write down is that critics had a positive view of the novel. You cannot write that the novel was a bestseller, that readers agreed, or that the author was previously unknown. Each of those would be a defensible real-world guess. None of them is forced by the stimulus, and only one answer choice will be forced.
The fourth move is to ask, of the answer choice, 'what would have to be true for this to be false?'. If the candidate can name even a single scenario consistent with the stimulus in which the choice fails, the choice is out. This is the single most reliable Must Be True elimination habit, and it scales. Once it becomes automatic, the test-maker's favourite traps become obvious. The next three sections dig into those traps in detail.
Trap 1: The imported word, especially the small adjectives
Most wrong answers on LSAT Must Be True questions are not wildly off. They are nearly right, and the wrongness is usually carried by a single word the stimulus never used. The candidate reads the choice, nods, and selects it. The word that did the damage is often a small modifier, a softened adverb, or a comparative that adds a dimension the stimulus did not mention. Examples that come up repeatedly: 'only', 'exactly', 'always', 'never', 'primarily', 'recently', 'best', 'most important', 'significant', and 'direct'.
Take a stimulus that says a city has increased the number of bike lanes and that cycling injuries per cyclist have fallen. A trap choice might read: 'The city now has the safest cycling infrastructure in the region.' That sentence uses 'safest' and 'in the region', neither of which appears in the stimulus. The defensible reading is that cycling has become safer per cyclist, not that the city leads its region. The correct answer would have to look something like: 'Per cyclist, the risk of injury on city streets has decreased.' Notice that the right answer mirrors the stimulus language almost exactly, while the wrong answer expands the claim.
For candidates still building a preparation plan, the practical drill is to write down the four most dangerous small words and check every answer choice for them. If a choice contains 'only', check that the stimulus actually justifies exclusivity. If a choice contains 'always', check that no exception is logically possible. If a choice contains 'best' or 'most', check that the stimulus ranks at all. If a choice contains a comparative like 'more than', check that the comparison is between the same two things. Roughly a third of missed Must Be True points in timed conditions come from a candidate who knew the right idea and selected the version with the extra word.
Trap 2: The softened quantifier that quietly reverses the answer
The second trap is the cousin of the imported word, and it deserves its own section because it does more damage per item. A Must Be True choice can be wrong not by adding new content but by softening a quantifier, which in turn reverses the direction of the conclusion. A stimulus might say that 'every member of the committee voted for the proposal'. A trap choice might say 'most members of the committee voted for the proposal'. Both are technically true given the stimulus, but only the first is forced. A candidate who marks the trap as correct has failed the stem's contract.
Another common softening pattern is moving from 'no' to 'few', from 'none' to 'some', or from 'all' to 'many'. The LSAT is unusually fond of these. The reasoning is simple: the test-maker can place a softened version of the right answer next to the strict version, and the test-taker who is moving quickly will pick the one that 'feels' more reasonable. On Must Be True items, reasonable is irrelevant. Forced is everything. A good rule of thumb: if the stimulus says 'all', the correct answer must also say 'all'. If the stimulus says 'some', the correct answer may say 'some' or 'at least one' but cannot be promoted to 'all' or 'most'.
In my experience, candidates often see a softened answer choice and select it because they assume the LSAT would not place a true-but-not-must-be-true option in the answer list. That assumption is wrong. The LSAT routinely places a true but non-forced choice alongside the correct one. Distinguishing them is the entire point of the question. A working preparation plan should include at least one timed set of twenty items where the only thing the candidate checks is quantifier alignment between stimulus and answer.
Trap 3: The right idea in the wrong direction
The third trap is the one that costs high-scoring candidates the most points, because the candidate sees the right concept and stops reading. A Must Be True choice can use the same vocabulary as the stimulus, name the same entities, and still reverse the direction of the relationship. A stimulus might say that 'the prevalence of the disease is higher in region A than in region B'. A trap choice might read 'region B has a higher prevalence of the disease than region A'. The candidate skims, recognises the names, and selects the trap.
Direction traps are especially common in causal stimuli. The LSAT will say that X leads to Y, and a choice will rephrase this as Y leads to X, or as Y causes X, or as the absence of X prevents Y. The same caution applies to comparative language. 'More than', 'less than', 'before', 'after', 'increase', 'decrease', 'weaken', 'strengthen' all carry direction. A Must Be True answer must preserve it. A useful habit is to draw an arrow over the relationship in the stimulus and then check that every answer choice points the same way.
For a candidate practising at home, a productive drill is to take ten Logical Reasoning stimuli, identify the directional claim in each, and then look only for direction in the answer choices. The first pass takes a while. The second pass is faster. By the third pass, the candidate starts noticing direction traps on the first read, which is where the time savings come from on test day. The LSAT Logical Reasoning section is not a speed test, but a section that runs out of time is a section that loses points, and direction-checking is a habit that costs almost no time once installed.
Trap 4: The conditional trap, including Must Be False look-alikes
The fourth trap is the conditional trap, and it sits at the boundary between Must Be True and a related question type that often appears in the same section. When a stimulus uses 'if', 'only if', 'unless', 'whenever', or 'in order to', it is offering a conditional structure, and Must Be True items will sometimes ask what must be true if the condition is met. The trap is the contrapositive confusion. 'If P then Q' guarantees that when P is true, Q is true. It does not guarantee that when Q is true, P is true. The test-maker knows this. A trap choice will assert the converse and rely on the candidate to misread the structure.
Within the same family sits the Must Be False look-alike, which sometimes appears inside a Must Be True set. The stem still says 'which one of the following must be true', but one of the choices describes a scenario that the stimulus rules out. The candidate who reads the choice as a statement rather than a check ends up marking the trap. The defence is mechanical. For each choice, the candidate asks: 'Is there a world in which the stimulus is true and this choice is false?' If yes, the choice is not forced and is therefore wrong. If no, the choice is forced and is the answer.
For candidates at the early stage of LSAT preparation, I usually recommend writing out the conditional in symbolic form on the scrap paper. P → Q, then ~Q → ~P is what the stimulus gives. Any answer that uses ~Q → P, or that treats P as equivalent to Q, is wrong on structure alone. By the time a candidate has done this for thirty stimuli, the contrapositive confusion stops appearing in wrong-answer analysis, and that is usually the moment their Logical Reasoning section score moves up a band.
How to prephrase the answer before looking at choices
Prephrasing is the single LSAT habit that converts a slow reader into a fast one, and on Must Be True items the technique is unusually powerful. The candidate reads the stimulus, identifies the load-bearing claims, and predicts, in plain English, what the answer must say. The prediction does not need to be grammatical. It can be a fragment. The point is to walk into the answer choices with a target shape. When a choice matches the shape, the candidate checks it. When a choice does not, the candidate moves on.
The technique works because Must Be True stimuli are short and the answer is usually a single sentence. The candidate has time, in most cases, to write a one-line prediction. In my experience, the best predictions focus on three elements: the quantifier, the direction, and the central entity. If the prediction says 'some of the X are Y' and a choice says 'most of the X are Y', the candidate has already identified the trap. The choice can be eliminated before reading the rest of the answer.
The discipline is to write the prediction before peeking. Candidates who glance at the answer choices before predicting almost always end up matching the closest-sounding choice rather than the forced one. The order matters: stimulus first, prediction second, choices third. The LSAT is a test of what the stimulus forces, and the test-maker has designed the choices to nudge a reader who skipped the prediction step.
Working through a Must Be True stimulus in real time
Consider a short stimulus: 'A bookstore stocks only paperbacks and hardcovers. Every paperback on the shelf was published after 2010. Some hardcovers were published before 2000.' A Must Be True stem asks which of the following must be true. The first move is to identify the conclusion, which in this case is descriptive rather than argumentative. The second move is to underline the quantifiers: 'only paperbacks and hardcovers', 'every paperback', 'some hardcovers'. The third move is to map the entities: the bookstore, the paperbacks, the hardcovers, the publication dates.
The prediction is: paperbacks on the shelf were all published after 2010. The hardcovers may include some pre-2000 titles, but the stimulus does not say all hardcovers are pre-2000. The bookstore contains only paperbacks and hardcovers, so no other format appears. With the prediction in place, the candidate scans the choices. 'Every paperback on the shelf was published after 2010' is forced and matches the prediction. 'No hardcover on the shelf was published before 2000' contradicts the stimulus. 'Most books on the shelf are paperbacks' is not forced. 'All books on the shelf were published after 2000' is not forced, because the pre-2000 hardcovers break it. Only the first choice survives the contract.
The same routine scales to longer stimuli. The candidate reads the conclusion, marks the quantifiers, draws the entities, predicts the answer, and then uses the four traps above to triage the choices. When the candidate is learning the routine, each stimulus takes longer than it will on test day. By the middle of a multi-week preparation plan, the same routine runs in well under the typical 90-second budget per Logical Reasoning question, and the candidate has the time to handle the few items that genuinely require careful re-reading.
Inference versus Must Be True: where the answer boundary actually sits
Many LSAT prep materials treat Inference and Must Be True as different question types. In practice, the two are scored and built the same way: the answer is forced by the stimulus. The label is a leftover from older test versions, and it sometimes confuses candidates into thinking a softer standard applies. On the current LSAT, the answer to an Inference question must be guaranteed by the stimulus, just as it would for a Must Be True item.
The boundary that actually matters is between Must Be True and a separate family of Logical Reasoning questions: the strengthen, weaken, and assumption items. Those questions ask the candidate to evaluate a claim, not to identify a claim that is forced. A candidate who blurs those two demands will mark an answer that strengthens the conclusion of a Must Be True stimulus, even when the stimulus does not force that strengthening. The trap is real, and it shows up most often in conditional stimuli, where an answer choice that the stimulus permits but does not require looks reasonable.
A practical distinction helps. For Must Be True and Inference items, the answer is the choice that survives every possible interpretation of the stimulus. For strengthen, weaken, and assumption items, the answer is the choice that interacts with the conclusion in a specific way. If a candidate reads a Must Be True stem and starts thinking about whether a choice helps or hurts the argument, the candidate has already misread the stem and should restart the question. The LSAT is unforgiving about this misread.
Common pitfalls and how to avoid them
The pitfalls below show up in nearly every candidate's first diagnostic. Treat them as a checklist to run after every timed set, not as a list to memorise once.
- Reading the answer choice that matches the candidate's personal view of the stimulus rather than the choice that the stimulus forces. The fix is to write a one-line prediction before peeking at the choices.
- Promoting a quantifier from 'some' to 'most' or 'all' inside the answer choice. The fix is to underline every quantifier in the stimulus and verify that each choice uses the same strength.
- Importing a small word the stimulus never used, such as 'only', 'always', 'best', or 'most important'. The fix is to keep a short list of dangerous small words and to check each choice against the list.
- Reversing the direction of a causal or comparative relationship. The fix is to draw a directional arrow over the relationship in the stimulus and to confirm that each choice points the same way.
- Misreading the stem as a strengthen or weaken question and selecting the choice that helps the argument. The fix is to re-read the stem after the prediction, before scanning the choices.
How Must Be True items fit into a broader LSAT preparation plan
A workable LSAT preparation plan treats Must Be True and Inference as a separate study strand rather than as a side effect of reading practice. The candidate allocates one weekly block to untimed stimulus reading, one block to timed sets of twenty items, and one short block to error review focused only on this question family. Over a multi-month plan, that structure produces a visible jump in Logical Reasoning section scores, because the candidate stops relearning the same habit with every section.
Untimed reading matters more than candidates expect. The goal in the untimed block is to install the four-move routine from earlier: conclusion, quantifiers, entities, prediction. The candidate reads each stimulus slowly, writes the prediction on paper, and only then opens the answer choices. Speed is not the goal. Accuracy under careful reading is the goal, because speed will follow once the routine is automatic. The timed block then converts the routine into a 90-second habit. The error-review block is where the real learning happens, and it should always end with a written note about which trap the candidate fell into.
For candidates who want a single number to track, the Logical Reasoning section is scored on a 120-180 scale, and a typical four-item swing in raw answers can move the score by a couple of points. The implication is that cleaning up Must Be True accuracy by a handful of items per section can move the overall LSAT score meaningfully. The same point applies to the Reading Comp section, where Must Be True stems appear in similar form, and the same trap inventory applies. Candidates who build the routine here will see spillover benefits in every other section of the test.
Conclusion and next steps
LSAT Must Be True and Inference questions reward a narrow habit set: read the conclusion, mark the quantifiers, map the entities, predict the answer, and then triage the choices for imported words, softened quantifiers, reversed directions, and conditional confusion. The habits are learnable, and they are transferable across the Logical Reasoning section and into Reading Comp. A candidate who builds the routine in untimed practice and then stress-tests it under timed conditions will see a steady climb in section scores, and the climb tends to show up in the overall LSAT score as well.
TestPrep İstanbul's diagnostic assessment is a natural starting point for candidates building a sharper preparation plan around Must Be True and Inference stems, because the diagnostic isolates the trap pattern that costs the most points in the first sitting and feeds it back as a targeted drill set.
Comparison: Must Be True, Inference, and Must Be False at a glance
The table below summarises the three related question types that often confuse candidates. Use it as a quick reference when reviewing error logs.
| Feature | Must Be True | Inference | Must Be False |
|---|---|---|---|
| Stem wording | 'must be true', 'logically follows' | 'can be inferred', 'follows logically' | 'must be false', 'logically ruled out' |
| Standard applied | Forced by the stimulus as written | Forced by the stimulus as written | Contradicted in every possible reading |
| Correct answer is | The choice the stimulus cannot contradict | The choice the stimulus cannot contradict | The choice that fails in every reading |
| Most common trap | Imported small word | Softened quantifier | Direction reversal |
| Defence habit | Write a prediction before peeking | Underline every quantifier | Try to falsify the choice |