LSAT Inference Question Tips

On LSAT inference questions, choose the answer that must be true given only the stimulus: the smallest claim you can prove from the text, not the most plausible one. Treat the passage as a closed universe: if you can't point to the lines that force an answer to be true, it's a trap.

What inference questions actually ask

Inference questions come in three strengths, and you should read the stem first so you know which standard applies:

• Must be true: "If the statements above are true, which one of the following must also be true?" The strictest standard: the answer is logically forced by the stimulus, with zero extra assumptions.
• Properly inferred / most strongly supported: "Which of the following can be properly inferred from the statements above?" Same skill, slightly looser bar: the answer should be strongly supported, not merely suggested.
• Cannot be true: the flip side: which answer is ruled out by the stimulus. If you can handle must-be-true, you can handle this by negating each choice.

You'll see several inference questions on every LSAT, mostly in Logical Reasoning but occasionally in Reading Comprehension. The core skill, separating what is proven from what is merely plausible, also powers your weaken, strengthen, and assumption questions, because all of them hinge on tracking exactly what the text does and doesn't establish.

The one rule: the text is a closed universe

Everything else follows from this. On inference questions:

• Facts in, assumptions out. You may only use what's stated in the stimulus and what logically follows from it.
• Your real-world knowledge doesn't count. If the stimulus says all employees at a firm take the bus, and an answer says bus ridership rose in the city, reject it, even if it's true in real life, the text doesn't support it.
• "Sounds reasonable" is not a standard. The correct answer is often the least dramatic one: a boring restatement, a combined fact, or a careful contrapositive. The wrong answers are usually the interesting, insightful ones.

This is the exact opposite instinct most students bring in. In everyday reading, you fill in gaps with common sense. On the LSAT, filling in gaps is the trap.

A 4-step must-be-true method

1. Read the stem first. Confirm you're in inference-land and note the standard (must vs. strongly supported vs. cannot be true). This tells you how strict to be before you read a single fact.

2. Map what the stimulus actually gives you. Many inference stimuli aren't even arguments: they're just bundles of facts. List what's there without adding anything:

• Separate facts from conclusions (and remember: even the "conclusion" in the stimulus counts as given, because the question asks what follows if it's all true).
• Watch for quantifiers: "all," "most," "some," "none". Conditional words: "if," "only if," "unless," "requires." These are the load-bearing words.

3. Look for what can combine. Valid inferences almost always come from combining two statements: rarely from a single line. The common patterns:

That third row matters: the LSAT loves giving you "most" and "some" statements that look combinable but aren't. "Most" plus "most" gives you nothing about the overlap unless you know the universe is exhausted. When in doubt, don't force the inference. The safest answer is usually a simple restatement.

4. Run the proof test on each answer. For each choice, ask: can I point to the exact lines that force this to be true? If you can't trace the chain, the answer is wrong. If two answers both look supportable, pick the one that says less: smaller claims are safer because they're easier to prove.

3 traps that catch almost everyone

1. The could-be-true answer. The classic wrong answer: a statement that's consistent with the stimulus and feels likely, but isn't forced. "The survey suggests customers prefer the new layout" → an answer saying "the new layout caused the preference change." Could be. Isn't proven. Wrong.

2. Real-world knowledge smuggled in. Answers that are true in the real world but unsupported by the text. This is why students who know the topic of a stimulus do worse on inference questions about it. They stop reading what's actually there. On inference questions, pretend the passage is about a fictional planet if you have to.

3. The strength mismatch. The answer over-claims: the stimulus says "some," the answer says "all." The stimulus says "may," the answer says "will." The stimulus says "often," the answer says "always." Any upgrade in strength needs proof: if the text didn't supply it, eliminate the choice. (Note the one-way street: softening a claim is fine, strengthening it is not.)

Inference vs. its neighbors

Students confuse inference with two related question types, and the distinction is worth getting straight:

• Inference vs. strengthen. Inference asks what is already proven by the text. Strengthen asks what new information would help the argument. On inference, a good answer never adds facts; on strengthen, it always does. If your chosen "inference" introduces something new, you've answered a different question.
• Inference vs. assumption. An assumption is an unstated piece the argument needs: you're finding what's missing. An inference is a stated-or-implied consequence: you're finding what's already there. If you find yourself reaching for conditional diagramming here, make sure you know which side of that line you're on: necessary vs. sufficient assumptions is where the diagramming reflex pays off most.
• Inference vs. flaw. A flaw question asks you to name what's wrong with the reasoning. An inference question asks you what's right: what the reasoning guarantees. If you're arguing with the stimulus ("but that doesn't prove it!"), you're in flaw-mode. Inference wants agreement: assume it's all true, then find what follows.