The NextGen bar exam reports exactly one number: a scaled score between 500 and 750. No MBE subscore, no written-day subscore — one number, with "Percent Below" context indicators alongside it that play no part in the pass decision. Whether you're admitted comes down to that number against your state's passing score.
Here's how the number is built, what's officially known versus honestly unknowable, and how to benchmark your practice against it without fooling yourself.
One integrated score, not three glued together
The legacy UBE added separately scaled components (MBE + MEE + MPT) into a 400-point total. The NextGen exam is scored as one integrated exam: every item — a 4-option MCQ, a select-two, a short written answer inside a question set, a 60-minute performance task — feeds a single scale.
By component, the composition is approximately:
| Component | Share of the score |
|---|---|
| Standalone multiple-choice questions | ~49% |
| Integrated question sets | ~21% |
| Performance tasks | ~30% |
The strategic headline hiding in that table: written and applied work carries about 51%. The MBE-era instinct — grind multiple choice, treat writing as a formality — is miscalibrated by half the exam.
Why "scaled" matters: equating across administrations
Your score isn't a raw percentage. NCBE equates scores across administrations so that a 615 in July 2026 represents the same demonstrated ability as a 615 in February 2028, even though the questions differ. That's what makes scores comparable across time and — critically — portable across NextGen jurisdictions, the way UBE scores moved between UBE states.
Equating has an honest consequence for candidates: nobody outside NCBE can tell you "X% raw = 615 scaled." Any prep resource that promises a precise raw-to-scaled conversion for a brand-new, equated exam is guessing. We prefer to say so plainly.
What your score report shows: one score, plus "Percent Below"
NCBE's testing column in the Spring 2026 Bar Examiner (Bob Schwartz) describes the report examinees receive: one overall scaled score on the 500–750 scale, accompanied by "Percent Below" values — the percentage of examinees who scored below you — at several levels: the total score, and groupings by question format and by subject area. Three cautions come with them. They are contextual indicators, not subscores, and are not used for licensure decisions. They should not be compared across administrations or read as trends. And their stated purpose is diagnostic: to help you identify where additional preparation would be most beneficial before a future attempt. Read that way, a low Percent Below on integrated question sets with a solid one on multiple choice is a February study plan, not a verdict.
The passing score: 610–620, state by state
NCBE recommends jurisdictions set their passing score between 610 and 620, describing that band as corresponding to the familiar 260–270 range on the legacy 400-point scale. Each jurisdiction chooses its own number. Among announced scores:
- 610 — Iowa, Missouri, New Mexico, North Dakota, Washington
- 612 — Guam, Palau
- 614 — Indiana
- 615 — Oregon (July 2026 only; 620 after), Northern Mariana Islands
- 616 — Connecticut, DC, Idaho, Kentucky, Maryland, US Virgin Islands
- 620 — Illinois, Maine, Nebraska, New Hampshire, North Carolina, Ohio, Pennsylvania, Vermont
Two quirks worth knowing: Oregon's 615 is explicitly a first-administration number that rises to 620 afterward, and Minnesota's court has guaranteed that any candidate scoring 620+ passes regardless of where its final cut lands.
Everyone else is "TBA" — in which case plan around the top of the recommended band. The always-current list, including UBE cut scores during the transition, is on the jurisdictions hub and in passing scores by state.
How written work gets graded
Multiple-choice items are machine-scored, including partial credit on select-two questions (one correct pick earns one of the two points; only both correct picks earn full credit — details in the question formats post).
Written responses — short answers inside question sets, performance-task products — are human-graded against rubrics, the way MEE and MPT answers were: per NCBE's Examinees' Guide, graders appointed by your jurisdiction score them on a common platform using NCBE's uniform rubrics and grader training, each response is dual-graded with reconciliation by grading leads, and responses are not graded or ranked relative to other examinees' work. The scoring guides, point scales and rubric levels are laid out in how written answers are graded. For your prep, that means practicing against a rubric matters: a written answer can feel complete while missing the two specific things graders award points for. BARGO's question sets and performance tasks include model answers and issue-by-issue self-grading rubrics precisely so your practice scores mean something.
Setting an honest practice benchmark
Since raw-to-scaled conversion is unknowable, what should you aim for in practice? Our approach, stated openly (and expanded week by week in the practice-scores readiness guide):
- We use 60% as the readiness benchmark on practice MCQs. The recommended passing band (610–620 out of 500–750) sits at 44–48% of the way up the scale; scoring comfortably above the midpoint of the scale range in practice — with honest, mixed-difficulty questions — gives headroom for exam-day variance. It's a heuristic, and we label it as one everywhere it appears.
- Estimates get guardrails. Our dashboard shows an estimated scaled score only after 50+ answered questions, flags that written components are self-assessed, and draws your state's actual passing line (or the NCBE midpoint, labeled as an estimate) rather than pretending precision.
- Trendlines beat snapshots. One quiz says little; accuracy over hundreds of questions, split by subject, says a lot. That's the number to move.
Start the trendline with the free 10-question diagnostic — it takes fifteen minutes and gives you a subject-by-subject baseline.
What we still don't know (and won't pretend to)
- Exact raw-to-scaled mapping — unknowable outside NCBE, by design of equating.
- Score release timing per state — set by each board of bar examiners; check yours.
- Most states' final passing scores — the TBA list is still long; NCBE's decisions page and our state pages track announcements.
The scoring system rewards exactly one thing you control: demonstrated ability across all three item types. Balance your practice the way the score is balanced — roughly half multiple choice, half written — and the number takes care of itself.
