.png)
LSAT flaw questions ask what is wrong with an argument's reasoning, not whether its conclusion is false. Recognize them from stems containing vulnerable to criticism, questionable because, or error in reasoning. Solve them by naming the gap in your own words before you read a single answer choice.
Recognition is the cheapest point on the section. Every flaw 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 youThe reasoning in the argument is most vulnerable to criticism on the grounds that itThe standard wording. Everything after that it is a candidate description of the flaw.The argument's reasoning is questionable because the argumentSame task, softer wording.Which one of the following most accurately describes a flaw in the argument's reasoning?Describes a flaw, not the flaw, so more than one real weakness may exist.The argument commits which one of the following errors of reasoning?Often signals a named, classic flaw.The argument is most vulnerable to the objection that it fails toThe fails to construction usually points at something the arguer overlooked.
Read the stem before the stimulus. Knowing you are hunting flaw 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 practiceFind the conclusion, then say the argument back to yourself as premise-therefore-conclusionYou cannot name a gap until you know which two things the gap sits between. Strip the sentence order the author chose and rebuild it as support and claim.Name the flaw out loud before you look at the answersThis is the whole technique. If you go to the answer choices without a prediction, you are shopping, and four of the five choices were written to sell you something. A rough prediction like it assumes the count and the rate are the same thing beats a polished one you found in an answer.Match your prediction to a choice, and treat abstraction as neutralCorrect flaw answers are often written abstractly, which makes them feel evasive. Read each choice as a description that either does or does not fit the argument you just rebuilt. If a choice describes a flaw that is real but is not in this argument, it is wrong.
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.
The city's bike-share program added 200 stations last year, and the number of cyclists injured in traffic rose by 18 percent over the same period. Clearly, bike-share stations make cycling more dangerous.
Question. The reasoning in the argument is most vulnerable to criticism on the grounds that it
(A) fails to consider that the number of people cycling may have risen along with the number of stations
(B) takes for granted that all injured cyclists were riding bike-share bicycles
(C) draws a conclusion about an entire city from evidence about a single neighborhood
(D) relies on a source with a financial interest in the program's expansion
(E) presumes without warrant that injuries are the only measure of danger
The answer is (A).
The conclusion is that the stations make cycling more dangerous, which is a claim about risk per cyclist. The evidence is a rise in the total count of injuries.
Those are different quantities, and the argument treats them as the same one. If the number of people cycling rose by more than 18 percent, cycling got safer while injuries went up.
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 failsBThe argument never claims anything about which bicycles the injured cyclists were riding, and it does not need to. Its conclusion is about cycling generally.CThere is no part-to-whole move here. The evidence is already citywide, so this describes a flaw the argument did not commit.DNo source is cited at all. This is a real flaw type that simply is not present, which is the most common shape of a wrong flaw answer.ETempting, because it sounds like a critique. But the argument does not need injuries to be the only measure of danger, only a measure of it, and that is uncontroversial.
Across flaw 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 real flaw that is not this argument's flawDescribes a genuine reasoning error, correctly worded, that this particular argument never commits. The most common wrong answer on the type.The premise objectionAttacks whether a premise is true. Flaw questions ask about the move from premise to conclusion, and premises are given.The overstated matchNearly your prediction, but with a word like only, always, or never that the argument never relies on.The counter-conclusionArgues the conclusion is wrong rather than describing why the support fails. Being wrong and being badly supported are different findings.
One question people always ask about flaw 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 flaw 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, flaw questions appear in both, and no single LR type is rare enough to skip or common enough to carry your score by itself.
Drilling flaw 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 flaw 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 flaw question you get wrong, write three lines before you look at any explanation: what you thought the flaw 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.
Look for vulnerable to criticism, questionable because, error in reasoning, or fails to. Each signals that your job is to describe what went wrong in the reasoning, not to evaluate whether the conclusion is true.
Rebuild the argument as premise-therefore-conclusion, name the gap in your own words before reading any choice, then match. Predicting first is what stops the wrong answers from working on you.
LSAC does not publish per-type counts, so no percentage is trustworthy. What is published is that two of three scored sections are Logical Reasoning, per LSAC's test specifications, and flaw questions appear throughout both.
No. Flaw answers describe a defect in the support, not a false conclusion. An argument can reach a true conclusion through terrible reasoning, and that argument is still flawed.
Learn the recurring shapes, but predict in plain language. Students who memorize labels tend to reach for whichever label sounds closest instead of checking whether the argument actually made that move.
Written by Ali, Georgetown Law, founder of Lovare Institut.
August 5, 2026
August 5, 2026