The LSAT is a strange test the first time you look at it. There's no history on it, no math, nothing you studied in college, and no body of knowledge to revise. It tests a small number of thinking skills, over and over, in two question formats. That's genuinely good news: a test of skills is a test you can get better at.
This guide covers what's actually on it: the sections, the scoring, and every kind of question you'll face.
The shape of the test
The LSAT is four 35-minute multiple-choice sections, with a 10-minute break between the second and third. Three of those sections count toward your score:
- Two Logical Reasoning sections — short arguments, one question each, roughly 25 questions per section.
- One Reading Comprehension section — four passage sets, roughly 27 questions.
The fourth section is an unscored variable section. It will be either another Logical Reasoning section or another Reading Comprehension section, it's used to test questions for future exams, and it is indistinguishable from the real thing while you're taking it. You will never know which section didn't count. LSAC doesn't tell you.
You cannot afford to write off a section you suspect is experimental. Some test takers convince themselves midway through that a hard section "must be the experimental one" and coast. Sometimes they're right. When they're wrong, it costs them several points on a section that counted.
There is also a separate writing task, LSAT Argumentative Writing, which is taken online on your own computer rather than as part of the test day sitting. It is not scored, but you must have a completed writing sample on file before LSAC will release your score.
Note that Logic Games, the section with the ordering and grouping puzzles formally called Analytical Reasoning, was removed in August 2024. If you find a study guide that still teaches them, it is out of date, and you should be suspicious of whatever else it tells you.
How the scoring works
You get one point for each correct answer across the three scored sections, for a raw score out of roughly 75 to 77. That raw score is converted onto the 120 to 180 scale, and the conversion shifts slightly from exam to exam to account for differences in difficulty.
Two consequences worth internalizing:
- There is no penalty for a wrong answer. A blank is worth exactly the same as a wrong guess: nothing. Never leave a question unanswered, and always leave yourself the last fifteen seconds of a section to fill in the ones you didn't reach.
- Every question is worth the same. The hardest Parallel Reasoning question on the test is worth exactly what the easiest Main Point question is worth. Spending four minutes on a monster while three easy questions go unread at the end of the section is a straightforwardly bad trade.
As a rough sense of scale, using LSAC's current percentile table: the midpoint of all test takers sits at about a 154, a 160 puts you in the top 30 percent, a 165 in the top 15 percent, and a 170 at roughly the 94th percentile. Percentile bands drift year to year, and they have drifted upward recently, so treat those as approximate and check the current figures against LSAC when you're setting a target.
Logical Reasoning: two-thirds of your score
Two of the three scored sections are Logical Reasoning: about 50 of the roughly 77 questions that count, or close to two-thirds of your score. If you have limited study time, this is where it goes.
Each question gives you a short passage of usually two to five sentences, called the stimulus, and asks one question about it. Most stimuli are arguments: a conclusion plus some evidence offered in its support. The arguments are almost always imperfect, in specific and repeatable ways, and nearly every question is asking you to do something with that imperfection.
Here is what one actually looks like:
Legislator: Under the new statute, a company is eligible for the small-business tax credit only if it has fewer than fifty employees. Dalton Tools has fewer than fifty employees. Therefore, Dalton Tools is eligible for the small-business tax credit.
Which one of the following most accurately describes a flaw in the argument's reasoning?
- It draws a conclusion about one particular company from a claim about small businesses generally.
- It treats a condition that is sufficient for eligibility as though it were necessary for eligibility.
- It treats a condition that is necessary for eligibility as though it were sufficient for eligibility.
- It relies on two different senses of the phrase "tax credit" at different points in the reasoning.
- It presupposes the truth of the very conclusion that it is trying to establish.
Show the answer
C is correct. The words that decide this question are "only if." Having fewer than fifty employees is something a company must have to qualify, but the statute never says it's the only requirement. Dalton Tools clears that one hurdle, and the argument concludes it has cleared them all.
B is the answer to beat: it correctly identifies that the argument confuses necessary and sufficient conditions, and then states the confusion backwards. Choosing between B and C takes about four seconds if you know the distinction cold and about ninety seconds if you don't, which is a fair summary of why that distinction is worth learning early.
That's the whole format. No outside knowledge, nothing to memorize, and a right answer that turns entirely on how carefully you read two words. Every Logical Reasoning question is some version of this.
LSAC doesn't publish an official list of question types, and prep companies carve them up differently. We track nineteen, which sort into four families by what they ask you to do:
Structural analysis
Describe what the argument is doing, without judging it. Includes Flaw in the Reasoning (name the error), Method of Reasoning (describe the strategy), Argument Part Identification and Role of Statement (what job does this sentence do), and Parallel Reasoning and Parallel Flaw (find the answer with matching structure).
Argument evaluation
Act on the argument. Strengthen and Weaken are the workhorses; Evaluate the Argument asks what information would help you judge it; Paradox gives you two facts in tension and asks what would explain both.
Assumption analysis
Find what's being taken for granted. Necessary Assumption asks what the argument requires; Sufficient Assumption asks what would be enough to make the conclusion follow. Those two sound alike and behave very differently, and the distinction is one of the most valuable things you can learn for this section. Principle questions move between a general rule and a specific case.
Deductive inference
Work out what follows. Must Be True and Most Strongly Supported ask what the statements guarantee; Main Point asks you to identify the conclusion; Conclusion / Fill-in-the-Blank asks you to supply it; Point at Issue asks what two speakers disagree about. These stimuli are often not arguments at all, just facts, and the discipline is staying strictly inside them.
Each type has its own characteristic trap: the wrong answer that a well-prepared student is most likely to take. Those are covered one by one in the full question type breakdown. And whatever the type, the method for working the question is the same three steps: analyze, predict, eliminate.
Reading Comprehension: four passages
One scored section, four passage sets, five to eight questions each. Most of the sets are a single passage of roughly 450 to 500 words. One set may be comparative reading: two shorter passages by different authors on a related topic, with questions about how they relate to each other.
That word may is new. From 2007 through 2025 every Reading Comprehension section had exactly three single passages and one comparative set. LSAC's current specification says a section contains either three or four single passages and either one comparative set or none, so as of the January 2026 exam you may sit a section without one. Prepare for comparative reading; just don't count on seeing it.
Subject matter rotates across law, humanities, social science, and natural science. It is deliberately unfamiliar, and it is not testing what you know about the topic. Everything you need is in the passage.
The questions come in three broad kinds:
- Structural reading — what the passage is doing overall. Main Point / Primary Purpose, Passage Organization, and Function questions asking why a particular detail or paragraph is there.
- Close reading — what it says. Specific Detail, Inference, Meaning in Context, and Analogy questions asking you to apply the passage's ideas to a new scenario.
- Perspective tracking — who thinks what. Author's Attitude, Author Would Most Likely Agree, Strengthen / Weaken applied to the passage's claims, and the Comparative questions about where two authors converge or diverge.
That third group is where a lot of points quietly go. LSAT passages routinely present two or three positions: the conventional view, a challenger, and the author's own take. The questions care enormously about who held which. Reading for content while ignoring attribution is the most common way to get Reading Comprehension wrong while feeling like you understood the passage perfectly.
There's more on how to read these passages in the Reading Comprehension guide.
The writing section
LSAT Argumentative Writing is done separately from your test day, online, with remote proctoring software. You're given a debatable issue and evidence on both sides, and you argue for one position.
It isn't scored, but every law school you apply to receives it, and your score won't be released until you've completed one. Practically speaking: don't spend real study time on it, don't skip it either, and write something organized and coherent. Admissions readers do occasionally look at it, particularly if your personal statement reads very differently from everything else in the file.
What this means for how you study
A few things follow directly from the structure of the test.
Logical Reasoning deserves the majority of your time. It's two of three scored sections, the question types are well-defined, and the improvement curve is steeper than in Reading Comprehension. It's also the skill that carries over, since getting better at finding a conclusion and spotting a gap makes Reading Comprehension easier too, since the passages are full of arguments.
Reading Comprehension is improvable, despite what people say. It's the section students most often write off as fixed. It isn't, but it doesn't improve by reading more carefully or more slowly. It improves when you change what you're reading for, from content to structure.
Track your accuracy by question type, not by section. "I got 71 percent on Logical Reasoning" isn't actionable. "I'm at 90 percent on Flaw and 45 percent on Necessary Assumption" tells you exactly what to do on Monday. Two people with identical section scores routinely need completely different study plans, and the only way to know which one you are is to count.
Timing is a symptom, not the disease. Nearly everyone runs out of time at first. That's usually because the underlying method isn't automatic yet, not because you need to read faster. Speed is what happens when the method stops requiring conscious effort.
If you're starting from scratch, the sequence that works is: learn what the question types ask, learn a repeatable method for working a question, drill untimed until it's second nature, and only then add the clock. That progression is laid out in how to improve your LSAT score.
Practice this with Forge180
Forge180 is built around the structure of the test: every Logical Reasoning and Reading Comprehension question type, tracked separately, so you always know which ones are costing you points.
- An AI tutor that coaches the analyze, predict, eliminate process instead of just showing you the answer
- Adaptive drills that pick each question from your actual strengths and weaknesses
- Analytics that show which question types and which skills are moving
Forge180 launches in August 2026. Email [email protected] to hear when it opens.