LSAT Parallel Reasoning Question Tips
Parallel reasoning questions ask one thing: which answer choice reasons the same way as the argument in the stimulus. The trap is built into the design: the test deliberately makes wrong answers about the same topic while the right answer is about something completely different. The fix: strip the content, match the structure (premises → reasoning move → conclusion), and eliminate choices that differ in conclusion type, reasoning type, validity, quantity words, or conclusion strength.
How to spot a parallel reasoning question
The stems are easy to recognize:
- "The reasoning in the argument above is most similar to which of the following?"
- "The pattern of reasoning is most closely paralleled by…"
- "Which of the following is most like the reasoning above?"
You'll typically see one or two per Logical Reasoning section, a few per test. Each one contains six full arguments (the stimulus plus five choices), which makes them the wordiest question on the test. That word count is exactly why they punish the wrong approach.
The two flavors: parallel reasoning and parallel flaw
There are two versions of this question type, and knowing which one you're facing changes your job:
- Standard parallel reasoning. The stimulus argument can be valid or flawed. Your job is to find the choice with the same reasoning pattern, whatever it is.
- Parallel flaw. The stem tells you the stimulus is flawed ("which of the following contains reasoning most similar to the flawed reasoning above?"). Your job: find the choice that commits the same error. This is where your flaw-spotting skills pay off directly: name the flaw in the stimulus first, then find the answer that makes the same mistake.
Both flavors reward the same method. Here's how it works.
The structure-first method
1. Read the stimulus and find the conclusion. Do this before you touch the answer choices. If you can't point to the conclusion, you can't abstract the structure.
2. Write the skeleton in plain words. Drop the nouns, keep the moves. "The council banned street parking to raise revenue, so the ban will increase traffic fines" becomes: They did X to get Y, therefore X will produce Y: a means-to-end prediction. It doesn't need symbols; it needs to be short enough to hold in your head.
3. Note the reasoning type and whether it's valid. Is it conditional ("if A then B")? Causal? An analogy? A statistical inference? And critically: is the reasoning good? A valid skeleton and a flawed skeleton are never parallel, no matter how similar the wording feels. For conditional arguments, make sure you're solid on necessary vs. sufficient conditions: confusing them is a common way to misread the skeleton.
4. Scan the answer choices' conclusions first, then eliminate. Don't read five full arguments. Read five conclusions and kill the ones that don't match your skeleton's conclusion type (a prediction about the future can't parallel a claim about what must be true). Then read the survivors' reasoning.
5. Confirm the winner fully. Elimination gets you to one or two candidates. Read the finalist completely and check every element of your skeleton is present. One missing premise disqualifies it.
The 5 elimination filters
Run every surviving choice through these filters. A mismatch on any single one kills the answer:
| Filter | What to compare | Example of a mismatch |
|---|---|---|
| Conclusion type | What kind of claim is it? (prediction, value judgment, recommendation, factual claim) | Stimulus predicts; choice recommends, not parallel |
| Reasoning type | Conditional, causal, analogy, statistical, sample-based | Stimulus is conditional ("if"); choice is causal ("because"), not parallel |
| Validity | Is the reasoning valid or flawed? | Stimulus commits a flaw; choice is airtight, not parallel, ever |
| Quantity words | All, most, some, many, few: these change the logic | Stimulus uses "most"; choice uses "all", not parallel |
| Conclusion strength | Must, will, probably, could | Stimulus concludes something might happen; choice says it must, not parallel |
The validity filter is the one most students skip, and it's the most powerful. When the stimulus is flawed, which it often is, you can eliminate every valid answer choice immediately. That's frequently three or four of the five.
The 4 traps LSAT writers set
1. The topic match trap. The stimulus is about electric cars and a wrong answer is also about electric cars. It's bait. LSAT writers build the credited answer about a deliberately unrelated topic. If a choice feels right because it's about the same thing, that's a warning sign, not a green light.
2. The language reuse trap. A wrong choice copies distinctive phrases from the stimulus ("just as the council failed to consider…"). Shared vocabulary ≠ shared structure.
3. The valid twin. Same reasoning type, same conclusion shape, but valid where the stimulus was flawed (or vice versa). This one catches careful students who did everything else right but didn't check validity.
4. The quantity-word shift. "Most" becomes "all," or "some" becomes "many." Small words, totally different logic. Read the quantifiers like they're the point of the question. Sometimes they are.
A quick worked example
Original argument (structure only, written for practice):
> All trail guides carry bear spray. Ren is a trail guide. Therefore, Ren carries bear spray.
Skeleton: All A are B. X is an A. Therefore X is B. Valid, conditional, universal ("all"), strong conclusion.
- (A) "Ren carries bear spray, so Ren must be a trail guide.": Reversed the conditional. Trap: same topic, wrong structure.
- (B) "All hiking boots need breaking in. These boots are hiking boots.": Same skeleton, but no conclusion stated. Missing element. Eliminated.
- (C) "Every certified lifeguard has passed the swim test. Dana is a certified lifeguard, so Dana has passed the swim test.": All A are B, X is A, therefore X is B. Different topic, same structure. This is the answer.
- (D) "Most trail guides carry bear spray. Ren is a trail guide, so Ren probably carries bear spray.": Weaker quantifier, weaker conclusion. Not the same logic.
Notice how (A) is the topic trap and (C) is the credited answer precisely because it isn't about the same thing.
Pacing: these are the slowest questions per point
Six arguments per question makes parallel reasoning the worst time investment in the LR section. Set a hard rule: if you haven't eliminated down to two choices within about two minutes, flag it, guess, and move on. Come back with time remaining. Losing one question to a flag costs you one question; burning four minutes on it costs you three.
Also watch the trend: when you're down to a coin flip between two choices, re-check the validity filter and the quantity words. Most coin flips resolve there.
Drill them right
- Do skeleton reps. Take five parallel reasoning questions and write only the skeleton for each stimulus, don't even read the choices. Ten minutes of this rewires how you read them.
- Study the near-misses. When you review a wrong answer, articulate exactly which filter should have killed it. "I picked (A) because it was about the same topic" is the single most valuable sentence you can write in your error log.
- Practice the flip. Time yourself on the elimination pass alone (stimulus → conclusions scan → two survivors). Speed on this pass is what converts these from time sinks to routine points.
Frequently asked questions
How many parallel reasoning questions are on the LSAT?
Typically one or two per Logical Reasoning section, so a few per test. They're rare enough that some students ignore them, but because each one is so mechanical once you learn the method, they're cheap points if you prepare.
Should I skip parallel reasoning questions?
Don't skip them permanently, but do flag-and-return when you're behind on time. Attempt them with fresh time: the elimination method is reliable, but it needs all your working memory. A rushed parallel reasoning attempt is worse than a guessed one.
What's the difference between parallel reasoning and parallel flaw?
In standard parallel reasoning, you match the reasoning pattern of the stimulus, whatever it is, valid or flawed. In parallel flaw, the stem tells you the stimulus reasoning is flawed, and you must find the answer choice that commits the same flaw. Name the error in the stimulus first; that's half the work.
Should I diagram parallel reasoning questions?
A short plain-English skeleton beats formal diagramming for most students. You need five things: the conclusion, the premises, the reasoning type, whether it's valid, and the quantity words. If conditional logic is involved and you want to diagram just that piece, our necessary/sufficient guide covers the notation.
Why do I keep picking the answer about the same topic?
Because the test is designed to make you. Topic similarity is the single most reliable trap on parallel reasoning: the writers deliberately match topics on wrong answers and mismatch them on the right one. Every time you catch yourself leaning on topic, re-read your skeleton and check the five filters instead.