What Inference Questions Actually Ask

Most Logical Reasoning questions hand you an argument and ask you to do something to it — find the assumption, expose the flaw, strengthen it. Inference questions are different in a way that trips up even strong test-takers: the stimulus usually isn't an argument at all. It's a pile of facts. The test isn't asking whether the reasoning is good. It's asking, given that everything here is true, what else has to be true?

That single shift changes your entire mindset. You are not a critic here. You are an accountant. You add up exactly what the stimulus gives you and you report the total — nothing more, nothing less. The correct answer is a statement you could prove using only the sentences in front of you. If defending an answer requires you to say "well, that's usually how it works" or "it makes sense that," you've already lost.

Must Be True vs. Most Strongly Supported

You'll see two flavors of this family. "Must be true" demands airtight proof — the answer follows with certainty. "Most strongly supported" is slightly looser: the answer is the one the stimulus makes most likely, even if it isn't logically guaranteed. The method is nearly identical, but on "most strongly supported" questions you tolerate a small inferential gap that would be fatal on a pure "must be true." When in doubt, treat the question as if it demands proof — you'll rarely be punished for being too strict.

What These Questions Test

Inference questions test a narrow but crucial skill: reading precisely and combining information without leaking in assumptions. The LSAT is rewarding your ability to distinguish between what is stated, what is proven, and what merely feels reasonable. Law school and legal practice live in exactly this gap — a fact in a record is not the same as a fact you'd like to be true. This is the LSAT training the muscle that keeps lawyers from overstating what their evidence actually establishes.

How to Recognize Them

The stem does the work. Look for language asking what the statements support or require, rather than what the argument does. Common phrasings:

• "If the statements above are true, which one of the following must also be true?"
• "The statements above, if true, most strongly support which one of the following?"
• "Which one of the following can be properly inferred from the passage above?"
• "Which one of the following is most strongly supported by the information above?"

The tell is the direction of support: the stimulus supports the answer. Contrast this with Necessary Assumption questions, where the answer supports the argument. Keep the arrow straight. On Inference questions, information flows down from the stimulus into the correct answer choice.

The Method

Run the same disciplined sequence every time.

1. Read for facts, not for an argument. Don't hunt for a conclusion. Inventory what you're told. Note any conditional statements (if/then), any quantifiers (all, some, most, none), and any hard commitments the author makes.
2. Look for connections. The single highest-yield move is spotting two statements that share a common term. If the stimulus says all A are B, and separately that all B are C, you can chain them: all A are C. That chained conclusion is very often the credited answer. Train yourself to ask, "What links to what?"
3. Don't pre-phrase a single answer. Unlike assumption or main-point questions, you usually can't predict the one specific inference the test will reward — a rich fact set supports many valid inferences. Instead, hold the facts firmly and go to the choices.
4. Test each choice against the standard: can I prove this? For "must be true," the question for every choice is literally, "Does the stimulus guarantee this?" If you can imagine a scenario consistent with the stimulus in which the choice is false, eliminate it.
5. Trust proof over plausibility. When a boring, obviously-supported answer sits next to an exciting, insightful one, the boring one usually wins. The LSAT rewards the safe inference, not the clever leap.

The Common Traps

The wrong answers here are engineered, not accidental. Learn the patterns and you'll eliminate faster than you can verify.

1. The Too-Strong Answer

The stimulus supports a modest claim; the answer states an extreme version. If the stimulus says something "can" cause an outcome, the trap says it "always" does. If the stimulus establishes a correlation among "some" cases, the trap says "most" or "all." Watch the quantifiers and modal words like a hawk. A single word of overreach kills a choice.

2. The Outside-Knowledge Answer

This answer is true in the real world — you know it's true — but the stimulus never establishes it. Your background knowledge is a liability on these questions. If proving the answer requires anything not on the page, it's wrong, no matter how obviously correct it seems.

3. The Reversed or Illegally-Negated Conditional

If the stimulus says "if X then Y," a trap will offer "if Y then X" (reversal) or "if not X then not Y" (illegal negation). Both are unsupported. The only valid contrapositive is "if not Y then not X." These traps punish sloppy conditional reading, so slow down when arrows are involved.

4. The Comparison That Wasn't Made

The stimulus discusses two things but never ranks them; the trap tells you one is greater, faster, or more important. If the text didn't establish a comparison, you can't infer one.

5. The New-Information Answer

The choice introduces a term, actor, or possibility that never appeared in the stimulus. A brand-new concept in an answer is a strong signal you can't prove it. Everything in a valid inference is built from parts the stimulus already gave you.

Worked Example

(Original example in the style of the LSAT — not a real LSAC question.)

Every registered vendor at the Harmon Market sells produce grown within the county. Some vendors at the market also sell baked goods. No vendor who sells baked goods is licensed to sell dairy.

If the statements above are true, which one of the following must also be true?

(A) Some vendors at the Harmon Market are not licensed to sell dairy.
(B) All vendors who sell produce grown within the county are registered at the Harmon Market.
(C) Most vendors at the Harmon Market sell baked goods.
(D) No registered vendor at the Harmon Market sells produce grown outside the county.
(E) Vendors licensed to sell dairy sell the highest-quality produce.

How to work it. Inventory the facts. Every registered vendor sells local produce. Some vendors sell baked goods. No baked-goods vendor is licensed for dairy. Now look for connections. "Some vendors sell baked goods" combines with "no baked-goods vendor is licensed for dairy" — chaining those gives us: some vendors are not licensed for dairy. That's answer (A). It's guaranteed.

Now the traps. (B) reverses the first sentence: we know registered vendors sell local produce, not that everyone selling local produce is registered here — classic conditional reversal. (C) turns "some" into "most" — a too-strong quantifier leap; "some" tells us nothing about "most." (D) sounds like a restatement of the first sentence, but it's actually a separate claim about all produce sold, and the stimulus only constrains registered vendors — it never says an unregistered vendor couldn't exist, so we can't guarantee this. (E) invents a comparison and a value judgment ("highest-quality") found nowhere in the text — pure outside information. The credited answer is (A).

A Few to Try

Original Lovare practice questions. Work them cold before reading the explanations.

How to Drill This Type

Inference questions reward a specific reading discipline, and the only way to build it is volume with ruthless review. Here's how to practice deliberately:

• Blind-review before you check answers. After an untimed set, revisit every question you weren't 100% sure on and force yourself to prove the answer from the text. If you can't articulate the proof, you got it right by feel — and feel fails on test day.
• For every wrong answer, name the trap. Was it too strong? Outside knowledge? A reversed conditional? Categorizing wrong answers turns a vague miss into a repeatable lesson. Over a few weeks you'll start pre-recognizing the pattern before you even finish reading the choice.
• Diagram the conditionals. When a stimulus is loaded with if/then and quantifier language, sketch it. The chaining move that yields the answer is far easier to see on paper than in your head.
• Slow down to speed up. Inference misses usually come from reading too fast and importing an assumption. Early in prep, take these untimed and prioritize precision. Speed comes from pattern recognition, not from rushing.

For scored, official practice under real conditions, use LawHub, LSAC's official platform — it contains real past PrepTests and is the authoritative source. Use Lovare's original drills to build the skill, then confirm it against official material. Remember that the current LSAT has two scored Logical Reasoning sections, so Inference questions carry real weight on your score — every point of accuracy here counts twice as much as it would have under an older format.