How to work each format · Published · Free, no account

How to work each question on the NextGen bar exam.

In the room, knowing what a format is won’t earn you much on its own.

You need to know which move to make first, what to do when two answer choices are still standing, and when to take the points you already have and go. That’s what we cover here.

But For · practice for the NextGen bar exam

Exhibit · The clockone mark per minute
Select-one2:00
Select-two2:30
Integrated set28 min
Standard task60 min
The question times are the targets we build to: most select-ones at two minutes, every select-two at two and a half. The set time is our suggested pace for its six parts. The task’s sixty minutes are NCBE’s.

The formats page sets out what each question type is and what it rewards. This one is the working manual for the same five formats: the order of the moves inside each, the time to give each question, and the decision to make when the clock is winning. It assumes you are learning the law somewhere else. Everything here is about what you do with it once a question is on the screen.

A word on where these come from. Some of what follows is arithmetic on the scoring, some is what published scoring guides reward, and some is a teaching judgment about what works, formed on this exam’s own question formats. Where it is our judgment, we say so rather than dressing it as a rule from the examiners.

One habit first: read the call before the facts

The call is the sentence that actually asks the question, and it usually sits at the end of the fact pattern. On the old exam you barely had to read it, because it nearly always asked the same thing: what is the legal result? On this exam it does not.

The same set of facts can carry a doctrine question, a research question, a counseling question, or a question about what a court already decided. Only the call tells you which one is in front of you, and each one makes a different fact the important one.

Watch what changes. If the call asks what the court will most likely hold, the elements of the rule are what you’re reading for. If it asks what to research next, the decisive fact is a hole in what you’ve been given, not a rule you know. If it asks what to advise the client, then what the client told you about her own situation becomes the operative law of the problem, and a technically sound answer that ignores it earns nothing.

So the order is: call, facts, then the call once more before you look at a single answer choice. Reading it twice costs about four seconds. Answering the wrong question costs the whole item, on law you may actually have known.

Select-one: four moves and a stop rule

One fact pattern, four answer choices, exactly one credited. The format is familiar and the work isn’t, because the call now asks what a lawyer would do at least as often as it asks what the law concludes.

Four moves, in order.

  1. Read the call and name the job. Doctrine, research, counseling, or what a court already did. Say which one to yourself before you read the facts.
  2. Answer the call in your own words before you read any answer choice. A prediction is the only real defense against an answer choice that is perfectly true and answers a different question.
  3. Read all four, and throw out on a named ground. A missing element, the wrong party, a rule that governs a different problem, or authority ranked in the wrong order. If you cannot name the ground, you have not eliminated it yet.
  4. If two answer choices survive, go back to the exact words of the question. A question asking what is most likely, one asking which fact best supports a position, and one asking what to do next can each have a different right answer out of the same four answer choices. The wording tells you which one it is asking for.

The stop rule. We build our questions of this kind for between ninety seconds and two minutes, most at two minutes. Those are our pace targets, not published ones, but the discipline behind them is worth borrowing. When two answer choices are still standing and you’ve already re-read the call, pick the one closer to the answer you predicted before reading the answer choices, mark it, and go. Whatever a third pass might add, it spends time you’ll want later on a question you can actually finish.

When the facts quote a statute, the answer turns on that text. This exam hands you law regularly, sometimes law you never studied, and it expects you to apply that text as far as it reaches, not the general rule you learned. Read the quoted passage before the answer choices and read it as though you knew nothing about the subject. The credited answer choice is the one that applies the text correctly, not the one that echoes its vocabulary, and a wrong answer choice can echo it perfectly. The habit of recognizing the doctrine and reaching for the majority rule is exactly the reflex the examiners built these questions to punish.

When you don’t recognize the law at all. This will happen, and the four moves above assume it won’t. All of them start from a prediction, and you can’t predict from nothing. Here is what to do instead, in order. First, read the call again and decide only what kind of answer it wants: a rule, a result, a next step, or a reason. That alone rules out answer choices of the wrong kind. Second, look for the answer choice that is doing something none of the others do. Three answer choices about the same thing and one about something else is a pattern worth noticing, and the odd one out is more often wrong than right. Third, throw out anything that just restates a fact from the question without adding a rule to it. Fourth, throw out anything that sounds fair rather than legal. Then pick from what is left, and go. A blind guess on four answer choices is worth a quarter of the credit on average, and every answer choice you remove pushes that up: three left is a third, two left is a half. There is no version of this where the blank is better.

Select-two: work it for points, not for the answer

Six lettered answer choices, A through F, exactly two credited, and partial credit that’s real. Get one right and you earn points. Get both and you earn more. A blank earns nothing at all, which makes leaving one blank the only choice on the entire exam that is guaranteed to be worth zero.

The arithmetic sits on the formats page, and it’s worth checking by hand once. What matters here is what the arithmetic tells you to do differently. Because this question type pays you for being half right, your job is no longer to find the answer. Your job is to turn whatever you actually know into as many points as it will buy, and then move on to the next question.

Five moves, in order.

  1. Read the call and decide what kind of pair it is. Is this two halves of one two-part test, or two independent reasons for the same position? You can usually tell from the call alone, and it narrows the field before you read an answer choice.
  2. Bank the one you are sure of. Going from nothing to one certain right answer is the single most valuable step on the question. Being sure of one answer choice is worth more than being half-persuaded by two.
  3. Name each answer choice that is left. True but beside the point, a rule for a different problem, the right doctrine one element short, authority in the wrong order, or the answer that just feels fair. Those five families are set out with examples on the formats page. An answer choice you can name is out.
  4. Check your two picks against the kind of question you decided it was. If it is a two-part test and both your picks argue the same part, one of them is wrong. If it wants two separate reasons and your two picks are near paraphrases of each other, same problem. Two answer choices that say the same thing are really one answer, and you need two.
  5. Move on when the time is up. We build ours for two and a half minutes. When the second correct answer won’t come, keep the one you’re sure of, take your best guess on the other, and go. Guessing the second one can’t cost you the point you already have, and the question you never reach scores exactly like a blank.

When three answer choices still look right. It’s almost always the same family doing it: statements that are perfectly accurate and answer a question nobody asked. Re-read the call and put one question to each survivor: does this answer what I was asked, or is it merely true? You rule out more answer choices that way than with any argument about doctrine.

When nothing looks right. Pick two anyway, and prefer the two that are at least about the right area of law. Even a pair chosen at random earns about a third of the question’s credit on average, and a pair chosen from the right area of law earns more than that. However lost you feel, a blank is never the better choice on this question type.

Integrated sets: answer the record in front of you

One client matter, one role, told in pieces. Documents arrive, questions follow, and new facts land partway through. Nothing closes behind you: the exam’s navigation pane lets you return to any question in the section, in any order. Our sets run about twenty-eight minutes across six components, which is roughly the pace to practice at.

The phrase that keeps appearing in the calls is the rule for the whole format: given the facts now known. Every component is answerable on what you have at that moment, and nothing revealed later makes an earlier answer wrong. That cuts both ways, because once you have more of the file you can’t go back and repair an answer with a fact you didn’t have.

Before you answer component one.

  • Say who you are and who your client is, in one sentence. You can only answer many of the client-advice questions if you remember what your role is, and you are told that role once, near the start, and never again.
  • Read the documents you were handed as facts, not as background. Interview notes are evidence. When a client says cash is tight, or that she will never use this vendor again, those sentences are the operative facts of every later counseling question.
  • Note the date and the posture. What has already happened, what has been filed, what deadline is running. Sets turn on timing more often than they announce it.

Do not hedge toward the twist. The instinct, once you realize facts are still coming, is to write an answer that covers the case you suspect is arriving. It reads as sophisticated and it scores badly, because it spends words on a record you don’t have yet and leaves the record you do have half-analyzed. Answer what’s in front of you, completely, and deal with the next component when it actually shows up.

Never leave a blank behind you. A blank earns nothing, and nothing on the screen will chase you about it. The exam’s navigation pane marks every question as answered, partially answered or unanswered, so open it while you still have minutes rather than at the end, when the clock is making your decisions for you. If you picked only one of the two right answers on a select-two, you still earn the point for the one you got, so a half-answer always beats a blank. It’s the cheapest loss to avoid in the whole format, and the easiest to miss.

When a question feels unanswerable. You reach a component and it seems to turn on something nobody has told you yet. Almost always you have misread it. NCBE’s instructions say plainly that no information later in a set is needed to answer an earlier component, so the fact you are waiting for is not coming, and everything above the question is already enough. Read the question again rather than reading ahead. This is the quiet time sink in the format: an unfolding matter feels like it is holding something back, so candidates stall on a question that was answerable when they got to it. And if you do move on, go back. The graders score the answer you leave, not the first one you gave.

Expect the subject to change without warning. A matter that opens in contracts can turn into an evidence question at component five, and nothing announces it. If you have only ever practiced one subject per session, this is the thing that will surprise you, and it is worth deliberately practicing sets that mix.

The standard task: how to spend the hour

A file, a library of authorities you’re given, one long writing assignment, sixty minutes. No outside law is wanted. The part people underrate is that the graders score the assigning memo: it tells you what to write, in what form, and what to leave alone, and each of those is worth points rather than politeness.

Here is the size of that. One of our own tasks runs twenty points, and two of them go to following the instructions and the format; on most of ours it’s one point in twenty. Either way, you settle that credit before a grader reads a word of your analysis.

Exhibit · The standard task · One way to spend the hourOur judgment, not a published rule
A minute-by-minute plan for a sixty-minute standard performance task
Minutes What you are doing What it buys
0 to 5 Read the assigning memo twice. Write its instructions down as the headings of your answer. The points for following instructions and using the right format, earned before you have opened a single source.
5 to 15 Read the file. Mark who wants what, and every fact with a date or a number on it. The specific facts you will build your elements from.
15 to 27 Read the law you were given, the cases and statutes, once through. In each one, mark the sentence that states the rule. A map you can return to without re-reading anything.
27 to 33 Outline under the headings you already wrote. One line per point, with its fact attached. You are writing now, not still deciding. Every remaining minute becomes words on the page.
33 to 55 Write, in the order of your outline, hardest issue first while you are freshest. The analysis points, which are most of the score.
55 to 60 Land a conclusion on every issue. Check the form against the memo one last time. Credit for stating a conclusion, plus the instruction-and-format points you secured in the first five minutes.

This split is a teaching judgment about how the hour is best spent, not a rule from the examiners. NCBE publishes the sixty minutes and the graded instructions; the boundaries above are ours, and they’re a starting point to adjust against your own timed runs. Don’t change the order, though: read and plan first, write second, and never leave an issue without a stated conclusion.

Read each source for one sentence. Most of the cases and statutes you’re handed carry a single line that states the governing rule, and that line is the one the scoring guide keys on. Find it, mark it, and move on. Later, when you need a case’s facts to compare against your client’s, the second read is fast because you already know where you are going.

The outline is the cheapest minutes on the clock. Six minutes of planning buys back more than six minutes of writing, because you end up rewriting unplanned prose, and rewriting in a document workspace on a clock is where hours go to die.

The research task: judging what you were handed

NCBE assembles the newer performance task differently: four multiple-choice questions, a short answer, and a medium-length writing assignment, all worked over a small library. What it tests is your judgment as a researcher, a skill the old exam never once made you show.

Before you answer anything, write one line about each source: what it is, and what rule it gives you. Four sources, four lines. It takes about five minutes and it answers most of the multiple choice on its own, because the questions are largely asking whether you did this.

  • 01
    What kind of authority is this?Primary or secondary. A statute, a regulation and a decided case are the law itself. A treatise, a practice guide and a law review article are somebody’s account of it, however distinguished the somebody.
  • 02
    Does it bind this court?Right jurisdiction, right level, still good law. An intermediate appellate decision from the correct state binds a trial court there and persuades everyone else. A sister state’s supreme court binds nobody here.
  • 03
    Is this the holding, or something said along the way?Find the question the court actually had to decide. A sentence you would love to quote, written about a situation the court was not deciding, is worth citing carefully and never worth relying on.
  • 04
    Does it reach my client’s facts at all?This is the question most candidates skip, and the one this task asks about most often. A source can be primary law, binding on your court, and squarely on your topic, and still never actually decide the exact situation your client is in.

That fourth line is the one worth practicing, because the credited answer is sometimes that your library doesn’t answer the client’s question. We built one of our research tasks exactly that way: four sources, and the correct response to the final multiple-choice component is that the controlling case expressly leaves this situation open.

Saying “the law I was given doesn’t settle this” is what a real lawyer does, and no exam has asked it of you before. Notice the trap in that task: if your written answer confidently announces a result the sources do not actually support, you have done the one thing the scoring guide exists to catch.

When two authorities pull in different directions. Ranking usually settles a research question: the statute beats the treatise, the appellate court beats the trial court. The harder case is two sources of comparable weight that do not agree, and this task asks about it more than candidates expect. One of ours turns on exactly this. A published appellate decision construes the statute’s key phrase but expressly leaves the client’s situation open. A recent practice commentary reports that trial courts have split on that open question. And the most tempting source in the library, the one that reads as if it settles everything, is not binding on this court. The move is to name the controlling source and follow it even when it says less than you want, then use the persuasive source only for what it is: a reason your court might lean one way, not an answer. Write the split down as a split. An answer that picks the more satisfying source and calls the question closed loses the point for the analysis and the point for candor. An answer that says “the controlling case does not reach this, the trial courts have divided, and here is which way the reasoning leans” earns both.

Written answers: writing to elements, not to length

Written work comes in two sizes, and the graders score both against a list of elements rather than an impression. The short one is the answer box inside sets and research tasks, which usually wants about a sentence. The long one is the analysis that closes a set or a task.

The one-sentence field

A field is a slot, and each slot pays once. Packing two facts into one sentence doesn’t earn two credits, and it risks earning neither if the grader can’t find a clean match. One fact per field, and make it specific enough to land on an element somebody wrote down in advance.

The difference is concrete. “The venue behaved unreasonably” isn’t an answer to anything. “The retained deposit equals forty percent of the contract price, far above the statute’s fifteen percent presumption threshold” is, because a scoring guide can find it.

The long analysis

Four things earn, and length isn’t one of them. State the governing standard in the words the provided text uses. Name the specific fact, with its number or its date. Spell out the comparison, which is the part graders actually award: this case is like that one, or unlike it, and here is the fact that makes the difference. Then land a conclusion, even a hedged one.

Here is a sentence from one of our own sample answers, on a prenuptial agreement task. The court is likely to hold that she signed voluntarily. She was told she could consult counsel, and the two-day window on its own isn’t enough to make the signing involuntary. Standard, fact, comparison, conclusion, in just over thirty words. “The agreement was signed voluntarily” reaches the same conclusion and earns a fraction of the credit.

A hedge isn’t a failure of nerve. An answer that ends with probably, and names the fact that makes it close, is scoring the conclusion point. An answer that simply stops, having analyzed well and decided nothing, isn’t.

And leave out what the task told you to leave out. The extra section on the issue the memo told you to skip earns exactly nothing and costs you the minutes you needed for the issue that counts.

When the clock is beating you

It will, at least once, on at least one section. Decide now, calmly, what you’ll do when it happens, because the exam room is a bad place to be working that out for the first time.

Steering the whole session, not just the question. Everything above tells you how long one question takes. Nobody answers just one question, though. You work three hours, and the questions don’t arrive evenly. So before you start, set yourself two checkpoints from the clock: where you should be at one hour in, and where at two. Write them down. If you are behind at the first checkpoint, do not speed up on the questions you are already working. Change what you skip. Mark and pass anything where two answer choices are still standing after your second read of the call, and come back to those last. If you are behind at the second checkpoint, protect the writing. Multiple choice takes the most time on this exam and performance tasks carry roughly a third of the score, so a written answer left blank costs more than a multiple-choice question guessed. Guess the multiple choice, in pairs where it is a select-two, and put your remaining minutes into the written work. Inside a task, spend minutes in proportion to points: our own tasks score their issues unevenly, six, eight and three, and the eight-point issue is not the one to leave for the last ten minutes.

Triage, in the order it pays.

  • Fill every select-two, first. Before anything else, go back and put two marks on any select-two you left open. The scoring table above is the reason: even a guessed pair averages a third of the credit, and a blank stays at zero.
  • Answer partially rather than not at all. The graders score written work element by element, so a paragraph that reaches two of four elements is worth two elements. A blank is worth nothing, and an unwritten conclusion is a point you had and did not take.
  • Take the conclusion before the flourish. If you are out of time mid-analysis, write the sentence that says where it lands. It is the fastest available point and the one people leave on the table most often.
  • Do not re-read what you have already answered. A second pass over a finished question feels productive, but whatever it might add, the unanswered question next to it is worth its full credit.

One thing worth knowing about yourself before exam day: whether your own sense of certainty means anything. If you’re usually right when you feel sure, you can move fast through those and spend the saved minutes on the ones you flagged. If your sure and your unsure are about equally accurate, that triage isn’t available to you, and every question gets the same second look whether it needs one or not. That’s a pacing problem long before it’s a knowledge problem, and it’s measurable in practice.

How to practice the moves

None of this survives one reading. These are habits, and you build habits by doing the thing on a clock and then finding out what it cost you. Two suggestions, and both are free to act on.

Practice each format the way it actually works, not by translating moves from an exam that no longer exists, and mix the sets across subjects: studying one subject at a time never trains you to make the switch. When you’re ready for a schedule, we publish the twelve-week plan in full, week by week, with an eight-week version for retakers. We answer shorter questions about any of this one at a time in the FAQ.

How we sourced this page

The exam facts here trace to NCBE’s published materials: the formats, the sixty-minute standard task, the graded instructions, and the partial credit on select-two. The pace figures describe our own questions and say so. The minute plan for the hour, the triage order, and the move-by-move sequences are a teaching judgment about what works, and this page marks them as ours rather than presenting them as rules from the examiners.

One honest limit. This exam has a single administration behind it, so nobody can tell you what worked for thousands of people last cycle, and anyone who claims otherwise is guessing at your expense. We reason here from the published scoring and from how the questions themselves are built. Where that reasoning turns out to be wrong, the correction will appear here and say what changed.

But For is an independent publication and study product of Ink Forms Rivers, d/b/a But For. We are not affiliated with, sponsored by, or endorsed by the National Conference of Bar Examiners or any bar examining authority; “NextGen UBE” is referenced solely to identify the examination. We publish examinee experience, never exam content.

support@butfor.ai reaches a human who reads every message, including about anything on this page you think is wrong.

What you can check

The .ai is for convenience. The vetting is not.

This page is about the moves. What goes wrong inside the law, subject by subject, is on where students go wrong. The Bench, our name for the practice app that comes with membership, is where you drill these moves until they’re habits. It holds 1,194 original questions across both multiple-choice formats, and eighteen integrated sets in which the facts arrive in order and nothing closes behind you. It also has both kinds of performance task against NCBE’s approximately 60-minute target, and written answers graded against a scoring guide you get to read. Fifteen questions are free to try, the grader will grade an answer you write, and we publish the editorial standard so you can hold us to it. Founding access is there when you’re convinced.