.png)
LSAT parallel reasoning questions ask which answer argues in the same shape as the stimulus, regardless of subject matter. Recognize them from most similar in its reasoning. Solve them by abstracting the stimulus into a structure first, because matching topic instead of structure is what makes these slow and wrong.
Recognition is the cheapest point on the section. Every parallel reasoning question announces itself in the stem, and the stems recycle a small number of formulations. Learn them and you stop spending seconds deciding what you are being asked.
Stem wordingWhat it tells youWhich one of the following arguments is most similar in its reasoning to the argument above?The canonical stem.The pattern of reasoning in which one of the following is most closely parallel to that in the argument above?Pattern of reasoning, same task.Which one of the following exhibits a pattern of flawed reasoning most similar to that in the argument above?Parallel flaw: now both arguments must be bad in the same way.The reasoning above is most closely paralleled by which one of the following?Compressed form of the same instruction.
Read the stem before the stimulus. Knowing you are hunting argument structure changes what you notice on the first read, and reading the stimulus cold means reading it twice.
Three steps, in this order. The order is the method, because the most common way to lose this question type is to read the answer choices before you have decided what you are looking for.
StepWhat it means in practiceAbstract before you read a single answerReduce the stimulus to a skeleton with letters. If X then Y, not Y, therefore not X. That skeleton is what you are matching, and writing it takes fifteen seconds you will get back four times over.Match validity first, then structure, then everything elseA valid argument can only be parallel to a valid argument, so a single check kills most choices immediately. Then match the number of premises and the logical form. Match the conclusion type last: a prediction pairs with a prediction, a recommendation with a recommendation.Ignore topic entirelyTest writers deliberately pair the stimulus with one answer on a similar subject and put the correct answer on an unrelated one. Subject-matter similarity is a decoy on this type more often than it is a signal.
The stimulus and answer choices below were written for this page. They are not taken, adapted or paraphrased from any real LSAT question, because reproducing licensed test content is not something we do. The style is the target; the content is ours.
Stimulus.
If the shipment had left the depot on Tuesday, it would have arrived at the warehouse by Friday. The shipment did not arrive at the warehouse by Friday. So the shipment did not leave the depot on Tuesday.
Question. Which one of the following arguments is most similar in its reasoning to the argument above?
(A) If the vaccination campaign had been effective, infection rates would have fallen. Infection rates did not fall. So the campaign was not effective
(B) If the vaccination campaign had been effective, infection rates would have fallen. The campaign was effective. So infection rates fell
(C) If the vaccination campaign had been effective, infection rates would have fallen. Infection rates fell. So the campaign was effective
(D) If the vaccination campaign had been effective, infection rates would have fallen. The campaign was not effective. So infection rates did not fall
(E) The vaccination campaign was effective only if infection rates fell. Infection rates fell. So the campaign was effective
The answer is (A).
The stimulus is if P then Q, not Q, therefore not P. That is a valid form.
Choice A is the identical form with different content: if effective then rates fall, rates did not fall, therefore not effective. The other four choices are the three classic conditional errors plus a rearranged conditional, which is what makes this set worth studying on its own.
Reviewing the four wrong answers is worth more than confirming the right one. Each was engineered to catch a specific misreading, and naming that misreading is what stops it recurring.
ChoiceWhy it failsBIf P then Q, P, therefore Q. Valid, but a different valid form, and parallel means the same form rather than merely also correct.CIf P then Q, Q, therefore P. Affirming the consequent, which is invalid, so it cannot parallel a valid argument.DIf P then Q, not P, therefore not Q. Denying the antecedent, also invalid.EOnly if introduces a different conditional relation, and the argument then affirms what is actually the necessary condition. Different structure and not valid.
Across parallel reasoning questions, the wrong answers fall into recurring shapes. Once you can name the shape you can eliminate on sight instead of arguing with each choice on its merits.
TrapHow it worksThe same-topic decoyShares the stimulus's subject and nothing else. Placed to catch anyone reading for content.The other valid formCorrect reasoning of a different shape. Valid is necessary but nowhere near sufficient.The conclusion-type mismatchMatches the premises but concludes a recommendation where the stimulus concluded a fact, or the reverse.The near-miss quantifierSwaps all for most or some, which changes the form even when the sentences look alike.
One question people always ask about parallel reasoning questions is how many appear per test. There is no honest answer to give. LSAC does not publish how many questions of each type appear on a given test, and no honest page can give you a percentage without inventing it.
What is published is the scored structure: the LSAT is two scored Logical Reasoning sections and one scored Reading Comprehension section, four 35-minute sections in total including one unscored variable section, with a 10-minute intermission after section two, per LSAC's own test specifications (https://www.lsac.org/lsat/register-lsat/accommodations/specifications-lsat-and-lsat-argumentative-writing). Two of the three scored sections are Logical Reasoning, so LR carries roughly two thirds of your scored questions.
That ratio is worth planning around. A per-type count is not, because nobody outside LSAC has one.
That matters practically, because study plans built on invented percentages allocate your weeks wrong. If you read that parallel reasoning questions are eleven percent of the test, ask where the number came from. It came from somebody counting a handful of released tests and rounding, which is not the same thing as a published specification.
Plan on the structure instead. Two of your three scored sections are Logical Reasoning, parallel reasoning questions appear in both, and no single LR type is rare enough to skip or common enough to carry your score by itself.
Drilling parallel reasoning questions in a block is efficient for learning the pattern and misleading about whether you have learned it. Blocked practice lets you know the question type before you read the stimulus, which is exactly the cue the real test removes.
The size of that gap is measured. In a classroom study of 787 students across 54 classrooms, interleaved practice produced 61 percent correct against 38 percent for blocked practice at a 33-day delay, an effect of d = 0.83, per Rohrer, Dedrick and Hartwig (2019), Journal of Educational Psychology (https://gwern.net/doc/psychology/spaced-repetition/2019-rohrer.pdf). Blocked practice felt better and tested worse.
The same reversal shows up in how students judge their own learning. Kornell and Bjork (2008), Psychological Science (https://gwern.net/doc/psychology/spaced-repetition/2008-kornell.pdf) found 78 percent of students judged massed study as good as or better than spaced study while 78 percent actually performed better spaced. Your sense that a drill block is working is not evidence that it is.
So use blocks to build the recognition, then stop. Once you can name a parallel reasoning question from the stem without thinking, the questions belong back in mixed sets where you have to identify the type yourself, because identifying it is half the skill.
Review is where the points are, and review is not reading the explanation for the answer you missed. The intervention research is blunt about this: across 28 studies and more than 16,000 participants, test prep as a category produced an overall effect of g = .26, and practice tests on their own produced g = .19 at p = .233, which is not statistically significant, per Hao and colleagues (2025), Review of Educational Research (https://journals.sagepub.com/doi/10.3102/00346543251360775). Taking questions is not the active ingredient.
What did work in that same Hao and colleagues analysis was structured practice on the reasoning: workbooks at g = .40 and explicitly teaching test-taking skills at g = .35. And there was no dose response at all, with neither contact hours at p = .489 nor program length in weeks at p = .314 predicting outcomes.
More hours is not the lever. Better review is.
Concretely, for every parallel reasoning question you get wrong, write three lines before you look at any explanation: what you thought the argument structure was, what it actually was, and the exact word in the stimulus that would have told you. The third line is the one that transfers.
Then leave the question alone for at least a day before you revisit it. Testing yourself beats restudying at g = 0.50 across 61 studies, the advantage rises to g = 0.69 when the delay is a day or more against g = 0.41 under a day, and feedback roughly doubles the effect at g = 0.73 with against g = 0.39 without, per Rowland (2014), Psychological Bulletin (https://pubmed.ncbi.nlm.nih.gov/25150680/).
Every LSAC figure on this page was taken from LSAC's own pages and last checked on July 26, 2026. Prices, dates and policies move without notice, so the source link sits next to each number rather than in a bibliography at the bottom.
Where LSAC has not published something, this page says so in those words rather than filling the hole with a plausible number. That is the difference between a gap and a guess, and only one of them is safe to plan around.
Most similar in its reasoning, or pattern of reasoning most closely parallel. If the stem adds flawed, both arguments must share the same defect.
Abstract the stimulus into letters before reading the choices, eliminate on validity, then match structure and conclusion type. Never match on subject matter.
LSAC publishes no per-type frequencies. The published structure is two scored Logical Reasoning sections and one Reading Comprehension section, per LSAC's own test specifications.
Only as a deliberate triage decision, and only if you are consistently out of time. They are long but mechanical, so they are often a better use of two minutes than a subtle assumption question.
Yes in the sense that a valid stimulus needs a valid answer and a flawed stimulus needs an answer flawed in the same way. Degree of persuasiveness beyond that is not the test.
Written by Ali, Georgetown Law, founder of Lovare Institut.
August 5, 2026
August 5, 2026