But For. The Report

The exam · A plain-English reference

What the NextGen bar exam actually is.

Nine hours over a day and a half. Three question families instead of the old components. Eight tested subjects, a great deal of law handed to you rather than memorized, and a score reported on a 500 to 750 scale. This is the structure, described without sales copy.

Almost everything written about the NextGen UBE was written by someone selling preparation for it, ourselves included. So this page holds to the rule the rest of the site holds to: structural claims trace to NCBE's published materials and official statements, and where we are reasoning rather than reporting, the sentence says so. NCBE PUBLISHED marks the former. ANALYSIS: OURS marks the latter. Nothing here reproduces exam content, and nothing here predicts what any administration will ask.

What it is, and what it replaced

The NextGen UBE is the licensing examination that began replacing the MBE-based Uniform Bar Exam with the July 2026 administration. NCBE PUBLISHED It is developed by the National Conference of Bar Examiners, adopted jurisdiction by jurisdiction on each jurisdiction's own schedule, and delivered on the examinee's own laptop through NCBE's secure platform at sites the jurisdiction manages.

The exam most candidates still picture has a multiple-choice component, a written component, and a performance test, each with its own name, its own day, and its own study strategy. The NextGen does not work that way. It runs as three question families that share a single platform and a single scaled score, and two of those three families ask you to write. The multiple-choice question survives, and it is still the largest single slice of the day, but it is no longer the exam's center of gravity in the way the MBE was.

The other structural difference is what you are expected to carry into the room. The legacy exam rewarded a very large memorized corpus. The NextGen tests a narrower list of subjects and then hands you a good deal of the governing law inside the questions themselves, expecting you to read it accurately and apply it as written. NCBE PUBLISHED That single design choice reorganizes almost everything about how the exam behaves.

Why it was rebuilt this way

NCBE does not leave the intent to inference. The exam names the foundational skills it means to measure, openly and in its own materials: legal research, legal writing, issue spotting and analysis, investigation and evaluation, counseling, negotiation and dispute resolution, and client relationship management. NCBE PUBLISHED Seven skills, and the new question types exist to reach them.

Read that list beside the old exam and the gap is obvious. A four-option multiple-choice bank can test issue spotting and analysis well. It cannot test whether you can rank a mandatory authority over a persuasive one, or whether you can weigh a settlement against a client's actual stated objectives rather than against the abstractly better legal outcome. Those are things practicing lawyers do constantly and the previous format never asked about. Weighing a settlement against a client's own interests has never been on a bar exam before. ANALYSIS: OURS

So the honest one-line answer to "why did they change it" is that the examination moved from testing what a new lawyer has memorized toward testing what a new lawyer does. Whether it succeeds at that is an empirical question that will take several administrations and real score data to answer, and anyone claiming to know now is guessing.

The shape of the exam

Nine hours, over a day and a half, in three three-hour sessions. NCBE PUBLISHED The exam is administered on examinee laptops through NCBE's secure platform, at testing locations arranged by each jurisdiction rather than by NCBE. That last detail sounds like logistics trivia. The first administration proved it is not, and we come back to it below.

Three item families carry the whole exam, and they divide the time and the score unevenly:

The three question families

Note the mismatch between those three labels. Multiple choice takes the largest share of the clock while performance tasks carry roughly a third of the score, which means the minutes and the points are not distributed the same way. Any study plan built on time spent rather than points available is optimizing the wrong quantity. ANALYSIS: OURS

What is tested, and what is not

The memorized syllabus is eight content areas: Contracts, Evidence, Civil Procedure, Business Associations, Real Property, Torts, Constitutional Law, and Criminal Law and Procedure, which NCBE treats as a single area rather than two. NCBE PUBLISHED That list is shorter than the corpus the legacy exam expected, and the shortening is the point rather than a rounding error.

Two further subjects are scheduled to enter. Family law and trusts and estates join the exam beginning with the July 2028 administration, largely through resource-provided formats rather than as new memorization. NCBE PUBLISHED

The narrowing happened inside the subjects too

The more useful fact for anyone planning a study cycle is that the content scope narrows within each subject, not only across the list. NCBE publishes a content scope outline per administration, and it is specific enough to read as a syllabus. When we mapped our own question bank against the outline governing the July 2026 through February 2027 administrations, doctrines a candidate would confidently expect turned out not to be listed at all. Under Evidence, the residual exception and the compromise-and-plea rules of 410 through 415 do not appear. Under Real Property, zoning and land use, equitable conversion, installment land contracts, and the Rule Against Perpetuities do not appear. Under Criminal, entrapment, insanity as a doctrine, and several whole crime families are absent. ANALYSIS: OURS, from mapping the published scope outline

Two cautions on that paragraph, and we would rather state them than have them found. First, the outline is republished per administration, so a topic absent from one scope is not permanently absent from the exam. Second, absence from the memorized outline is not the same as absence from the exam: an unlisted topic can still appear if the question supplies the law you need. Treat the current outline as the syllabus it is, read it directly, and check the edition against the administration you are registered for.

Provided law, the deepest change

If you take one structural fact from this page, take this one. The NextGen regularly hands you a statute, a rule, or an edited opinion and expects you to apply that text, including where it departs from the general rule you memorized. It can even reach subjects outside the memorized syllabus entirely, supplying all the law required. NCBE PUBLISHED The criminal outline is explicit on the point: elements of crimes and defenses are tested on the basis of provided statutes. NCBE PUBLISHED

Provided law sounds like a mercy and is closer to a trap. A well-drilled candidate's fastest instinct is recall, and recall is exactly what a modified statute punishes. The skill being tested is close reading against a trained reflex, which is harder under time pressure than either reading or recall alone. It is also, plainly, the thing lawyers do: nobody in practice applies the majority rule they memorized in law school without first checking what their jurisdiction's text says. ANALYSIS: OURS

How it is scored

The NextGen reports on a 500 to 750 scale, with the passing line set by each jurisdiction rather than nationally. NCBE PUBLISHED Results historically arrive months after an administration, not weeks.

Below the scaled score, two mechanics are worth knowing before exam day. Select-two questions carry genuine partial credit, two points, one point, or none, so guessing dominates abstaining on every single one of them. And on the standard performance task, the formatting and the instruction-following are themselves graded, which means an answer that is legally correct and structurally noncompliant leaves points on the table.

What nobody has yet is calibration. Because the first jurisdictions to adopt were small, and because pass lines are jurisdiction-set decisions that arrive on their own timetables, there is no meaningful public difficulty baseline for this exam. Any provider quoting you a difficulty rating, a predicted score band, or a pass-rate expectation for the NextGen is describing a feeling, not a measurement. That is also why our own score bands and difficulty estimates stay labeled provisional until real data earns the label's removal. ANALYSIS: OURS

When it reaches each jurisdiction

Adoption is jurisdiction by jurisdiction, in four waves. The dataset below traces to NCBE's published adoption list and was verified in August 2026. NCBE PUBLISHED Jurisdictions announce and occasionally revise on their own authority, so confirm your own against your bar admissions agency's direct communications before you build a calendar on it.

July 2026

The first administration. Connecticut, Guam, Idaho, Maryland, Missouri, the Northern Mariana Islands, Oregon, Palau, the U.S. Virgin Islands, and Washington.

Ten jurisdictions, roughly 2,624 examinees scheduled across sixteen testing locations.

July 2027

Thirteen more join. Arizona, Iowa, Kentucky, Minnesota, Nebraska, New Mexico, North Dakota, Oklahoma, South Dakota, Tennessee, Vermont, West Virginia, and Wyoming.

February 2028

Delaware, the District of Columbia, Illinois, and Montana switch.

This is the legacy UBE's final administration.

July 2028

The large wave. Alabama, Alaska, California, Colorado, Florida, Georgia, Hawaii, Indiana, Kansas, Maine, Massachusetts, Michigan, New Hampshire, New Jersey, New York, North Carolina, Ohio, Pennsylvania, Rhode Island, South Carolina, Texas, Utah, Virginia, and Wisconsin.

With California, New York, Texas, and Florida in, the majority of American bar candidates are now taking the NextGen. California's added state-specific component arrives no earlier than July 2029, per the California Supreme Court's July 2026 order.

Unannounced

As of August 2026, Arkansas, Louisiana, Nevada, and Puerto Rico had not announced a date. Mississippi had signaled adoption without one.

The practical reading of that table: if your exam falls at or after your jurisdiction's switch date, you are taking the NextGen, and materials built for the MBE describe an examination you will never see.

What the first administration showed

Approximately 2,624 applicants were scheduled across ten jurisdictions and sixteen testing locations for the inaugural administration in July 2026. NCBE PUBLISHED Seven of the ten appear to have administered without reported incident. Two, Maryland and Missouri, were disrupted and completed the same day. One did not run at all: Washington canceled its entire July exam after an on-site failure at its testing venue, without a single question answered by its standard-time examinees.

The lesson relevant to this page is narrow and structural. NCBE's position is that the week's failures were site-local, involving venue networks, on-site equipment, and local configuration, and did not involve the testing platform itself. NCBE PUBLISHED The evidence is consistent with that. It is also cold comfort in the room: an exam delivered on personal laptops in jurisdiction-managed venues carries a delivery risk the paper era did not, and preparing for this exam now sensibly includes knowing your own jurisdiction's contingency policy. The complete sourced account, with every claim labeled, is the first-administration report.

What to understand before preparing

Five things worth settling before you spend a dollar or an hour.

Confirm your administration first. Everything else follows from whether your exam falls before or after your jurisdiction's switch date. Confirm it with the bar admissions agency directly, not with a prep company's chart, this one included.

Read the content scope for your administration. It is published, it is specific, and it is shorter than the syllabus you have been told to expect. Reading it yourself is the highest-yield hour in the entire cycle.

Outlines and drilling are different jobs. Knowing the rule and executing the format under a clock are separate capacities, and the second one is what the new question families measure. Whatever you use for the first, the second needs practice in the exam's own grammar.

Distrust confident numbers. Difficulty ratings, predicted score bands, and pass-rate promises for an exam with almost no public score history are not available to anyone yet, whoever is selling them.

Practice the writing, not just the reading. Two of the three question families require written work, and the performance tasks alone carry roughly a third of the score. Written answers that are never read by anyone, or never graded against a rubric, do not improve.

Questions, answered plainly

How long is the exam?

Nine hours, over a day and a half, in three three-hour sessions, on your own laptop through NCBE's secure platform at a site your jurisdiction manages.

Is it harder than the old bar exam?

Unknown, and anyone who tells you otherwise is guessing. There is no prior NextGen administration to compare against and no public score history, so "harder" currently measures a candidate's expectations, set largely by prep materials, rather than the exam against its own baseline.

Do I still have to memorize the law?

Yes, for the eight tested content areas, at the depth the published scope outline specifies. What changed is that a great deal of additional law arrives inside the questions, and applying provided text accurately, including where it conflicts with what you memorized, is itself a graded skill.

Can I use old MBE question banks?

They still exercise doctrinal recall, and recall still matters. What they cannot exercise is the select-two format and its partial credit, the integrated sets and their unfolding records, or the legal-research task, because none of those existed. A candidate relying entirely on legacy materials has practiced roughly one of the three families.

What happens if my testing site fails?

That is a jurisdiction question, not an NCBE question, and the answer differs by jurisdiction. The first administration produced one statewide cancellation with a three-way remedy offered to affected applicants. If it matters to you, ask your admissions agency what its contingency policy is before exam day rather than after.


But For is an independent publication and study product of Ink Forms Rivers, d/b/a But For, and is 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. Structural and rollout information on this page traces to NCBE's published materials and official statements and is provided for orientation only; registered applicants should rely on their own jurisdiction's direct communications and on NCBE's current published materials for anything they act on. We publish examinee experience, never exam content.

Everything above is the exam. But For is the bench built to it, format for format: 534 original questions including 216 select-twos with the real partial credit, twelve integrated sets, all six performance tasks in both formats, and AI grading that quotes its evidence. The three-question calibration is free, the editorial standard is published, and founding access is there if the evidence convinces you.