Rank 1 by frequency | 668 questions in corpus (14.8% of all questions)
Flaw questions present an argument containing a reasoning error and ask you to name the error — not fix it, not replace it, just describe what is wrong with the logic as it stands. This is the single most common question type on LSAT Logical Reasoning, and the correct answer is always an abstract description of the logical mistake the argument commits.
Flaw is a descriptive task. You are identifying why the conclusion does not actually follow from the premises — why the evidence the author offers is insufficient, the wrong type, or depends on an unjustified leap. The correct answer characterizes the mistake in abstract, logical terms; it does not provide new information, supply the missing premise, or offer a better argument.
In every Flaw stimulus, the conclusion outstrips what the premises can logically establish. That overreach boils down to one of three things: the reliability of the evidence is insufficient, the terms of the conclusion exceed what the evidence establishes, or a hidden assumption bridging premise to conclusion is unjustified. Your job is to spot which of those gaps the argument walks into and match it to LSAC's abstract phrasing.
Because this is a descriptive task, every correct answer must survive two simultaneous tests: does the argument actually do this? and is this actually a problem? An answer that describes a real flaw the argument doesn't commit fails the first test; an answer that describes something the argument does but that isn't logically troublesome fails the second.
Before you can name a flaw, you need a vocabulary of flaw types. LSAC draws from roughly 27 categories, but a much smaller group accounts for the majority of questions. Learning these patterns is the single biggest unlock for this question type.
Causal/correlation confusion. The argument spots a correlation between two things and concludes one caused the other. Sub-patterns include confusing correlation with causation, post hoc reasoning (temporal sequence treated as causation), ignoring reverse causation, ignoring a common third cause, and treating coincidence as a causal link. Roughly every time the LSAT concludes A causes B from correlation evidence, this flaw is present.
Conditional reasoning flaws (necessary/sufficient confusion). The argument misapplies a conditional statement — either through a mistaken reversal (affirming the consequent: A→B, B, therefore A) or a mistaken negation (denying the antecedent: A→B, not A, therefore not B). These are logical equivalents of each other and are among the most tested flaws on the LSAT.
Circular reasoning. The conclusion is supported by a premise that is essentially the same claim restated in different words. The argument assumes what it is trying to prove. Easy to spot when obvious, very hard when the two versions use different vocabulary.
Ad hominem / source argument. The argument rejects a claim by attacking the person making it rather than addressing the claim itself. Sub-patterns include attacking character, attacking past acts (tu quoque), attacking motivation or bias, and attacking the origin of the idea. An argument's logical merit is independent of who makes it.
Equivocation. A term with multiple meanings is used inconsistently throughout the argument — for example, happiness meaning both "philosophical self-actualization" and "casual pleasure." The semantic shift is often extremely subtle, which makes equivocation one of the higher-difficulty flaw patterns.
Other common patterns. Sampling and representativeness errors, numbers-vs-percentages confusion, composition/division (part-to-whole and whole-to-part), appeal to ignorance, false dichotomy, and scope/strength overreach all show up regularly. Together with the five above, these account for the vast majority of Flaw answers.
Flaw stems are relatively uniform. Most ask some version of "the argument's reasoning is flawed because..." or "the reasoning is vulnerable to criticism on the grounds that..." or "which one of the following most accurately describes a flaw in the argument." A small number add wrinkles: Parallel Flaw questions ask you to find a second argument with the same flaw as the first, and Dialogue-format flaw questions present two speakers and ask about the flaw in one speaker's reasoning. The underlying task is identical — identify and describe the logical error.
Step 1 — Identify the conclusion. Look for conclusion indicators (therefore, thus, hence, consequently, so) or identify which claim everything else is trying to support. You cannot spot the flaw until you know what the author is trying to prove.
Step 2 — Identify the premises. Separate the actual evidence from background context, counterarguments, and concessions. Only the premises the author offers in support of the conclusion matter for the flaw.
Step 3 — Ask "why don't these premises prove this conclusion?" Resist the urge to subconsciously help the argument. Your job is to be skeptical, not charitable. Where does the logic break down?
Step 4 — Prephrase the flaw. Before looking at the answers, put the error in your own words — "the author assumes correlation means causation" or "the author treats a necessary condition as sufficient." A concrete prephrase makes matching abstract answer-choice language much faster.
Step 5 — Use the Mad Libs technique. Fill in the abstract terms in each answer choice with the concrete terms from the stimulus. If the filled-in description accurately maps onto what the argument does AND describes a real problem, you've found your answer.
Step 6 — Apply both filters to each choice. For every candidate, ask: (a) does the argument actually do this, and (b) is this actually a problem? One wrong clause invalidates the entire answer.
Step 7 — If stuck, use Plan B. When you can't articulate the abstract flaw, think about how you would weaken the argument. If you can see how to damage it, you fundamentally grasp the flaw — then find the answer choice that describes that weakness as a logical error.
Correct Flaw answers are abstract and descriptive — unlike a Weaken answer, which is concrete and factual, a Flaw answer characterizes a reasoning pattern using abstract variables the reader must map onto the specific argument.
Two phrases dominate the correct-answer language. "Takes for granted that..." and "presumes, without providing justification, that..." signal that something was improperly included — the argument relied on an unjustified claim. "Overlooks the possibility that..." and "fails to consider that..." signal that something was improperly excluded — the argument ignored a relevant alternative. These two families are virtual opposites, and recognizing which one applies is a fast way to narrow the field.
Other frequent phrasings include "confuses X with Y," "treats a necessary condition as sufficient," "generalizes from an unrepresentative sample," and "mistakes a correlation for a causal relationship." Building familiarity with this vocabulary is half the battle on Flaw questions.
Trap 1 — Describes something the argument does not do. The most common trap. Sounds like a real flaw type but the argument doesn't actually commit that error. Defense: apply the Mad Libs technique and verify the description maps onto the specific argument.
Trap 2 — Describes something that is not actually a flaw. Accurately describes part of the argument, but the described feature is not a reasoning error. The argument does this, but doing this isn't a problem.
Trap 3 — Right flaw, wrong scope. Similar to the actual flaw but too narrow or too broad — misses the specific shape of what went wrong.
Trap 4 — Right concept, wrong wording. Has the correct general idea but reverses which element is treated as which (for example, confuses necessary with sufficient by putting the pieces in the wrong order).
Trap 5 — Partially accurate. Describes part of the argument correctly but fails to capture the actual error. One wrong clause invalidates the whole answer.
Trap 6 — Irrelevant flaw. A genuine, named flaw type that simply doesn't apply to this particular argument. Easy to fall for if you've been primed to spot common flaws without checking fit.
Abstract answer-choice language is the biggest difficulty multiplier. When every choice is wrapped in dense logical terminology, the correct answer can feel indistinguishable from wrong ones until you apply Mad Libs carefully. Subtle shifts in strength, scope, or meaning between premises and conclusion — nearly invisible word-level changes — also drive difficulty up.
Other difficulty levers: arguments with multiple apparent flaws where only one is tested, unstated assumptions as the flaw (where the flaw is the hidden assumption), and dense or confusingly worded stimuli where just identifying the conclusion takes careful reading. On the hardest versions, LSAC stacks several of these features at once.
For practice, several classic examples illustrate the range: PT20 S4 Q14 (Difficulty 2, causal fallacy) concludes owning a laptop led to a higher-paying job from a 25% earnings correlation. PT22 S4 Q21 (Difficulty 4, necessary/sufficient) features Terry and Pat committing parallel errors about "favorable consequences" and goodness. PT51 S1 Q6 (Difficulty 3, equivocation) has Deirdre shifting between two meanings of "happiness." PT14 S4 Q15 (Difficulty 4, ad hominem) dismisses the Environmental Commissioner's proposals because they resemble those of a polluter.
vs. Weaken. Both require identifying a weakness in the argument, but Flaw asks you to DESCRIBE the error in abstract terms while Weaken asks you to PROVIDE new information that damages the argument. Flaw answers characterize; Weaken answers introduce facts.
vs. Method of Reasoning. Both describe how an argument works in abstract terms, but Flaw focuses on what goes WRONG; Method describes the reasoning pattern neutrally, whether or not it is flawed.
vs. Necessary Assumption. The flaw IS the unjustified assumption, just viewed from a different angle: Flaw asks "what error did the author commit?" while Necessary Assumption asks "what must the author be assuming?"
vs. Match the Flaw. Match the Flaw requires identifying the flaw AND finding a second argument with the same flaw. Standard Flaw stops at identification.
Flaw stems almost always include some form of the word flaw, flawed, vulnerable, criticism, or error. Recognizing them should immediately trigger the conclusion-premise-gap workflow and activate your mental library of common flaw patterns.