How NP Certification Exams Are Scored: Scaled Scores, Pass Marks and What They Mean

NP certification exams are pass or fail, and the pass mark is a fixed standard rather than a ranking against other candidates. ANCC states that an examinee "must achieve a scaled score of 350 or higher" on a scale whose maximum is 500, after the raw number of correct answers is converted by a formula. NPCB (the American Academy of Nurse Practitioners Certification Board, doing business as the Nurse Practitioners Certification Board) sets its passing score "using recognized standard-setting methodology" and states plainly that scores are not reported as a percentage. Understanding what sits behind those sentences removes most of the anxiety about how many questions can be missed.

Raw score, scaled score, pass mark

Three numbers are involved, and only one is yours to control.

• Raw score. The count of scored items answered correctly. ANCC's handbook uses "105 out of 150" as an illustration of a raw score, not as a pass mark. On NPCB the raw score is out of 135.
• Scaled score. A transformation of the raw score onto a fixed reporting scale so that results from different exam forms mean the same thing. ANCC's scale tops out at 500. NPCB's current handbook does not name a scale; the board's scoring page has historically described a 200 to 800 scale with 500 required to pass, and the 2026 handbook says only that scores are not reported as a percentage, so treat the specific range as something to confirm in your own score letter.
• Pass mark (cut score). The minimum performance that a panel of practicing NPs judged necessary for safe entry-level practice. It is set once per exam version, then applied to every candidate.

The distinction between raw and scaled matters because exam forms differ slightly in difficulty. If one form happens to contain a few harder items, the raw score needed to reach the same scaled score drops by the same small amount. Equating protects candidates from the luck of the draw; it also means no one can state a universal "number you can miss."

How the pass mark is set

Both boards describe criterion-referenced standard setting. ANCC names the method: the modified Angoff procedure. In an Angoff study, a panel of subject-matter experts, guided by a psychometrician, imagines a "minimally competent" new NP, the provider who is just safe enough to practice unsupervised. For each item the panelists estimate what proportion of such borderline candidates would answer correctly. The estimates are averaged across items and panelists, and the sum becomes the raw cut score for that form. The cut score is then mapped onto the reporting scale; ANCC maps it to 350.

NPCB describes the same philosophy in general terms: a passing score "established using recognized standard-setting methodology" that reflects "the knowledge required for competent nurse practitioner practice." The practical consequence is identical. The standard does not move with the ability of the people in the room on a given day, and a strong cohort does not push anyone below the line.

Why pretest items exist and why they cannot be spotted

Every exam form carries unscored items: 15 of 150 on NPCB, 25 of 175 on ANCC. The boards use candidate performance on these items to decide whether they are fair, discriminating and correctly keyed before they are promoted to scored status on a future form. The ANCC outline states that pretest items "cannot be distinguished from those that will be scored." Two habits follow from this:

• Answer every question with full effort. A strange or seemingly out-of-scope item may be scored; a routine one may be the pretest.
• Do not let an item you are sure is "experimental" change your pacing. There is no reliable tell.

Pretest performance never raises or lowers a score. A candidate who finds 20 items baffling on a 150-item NPCB exam has still faced at most 15 unscored ones, so the arithmetic of a borderline result is not as forgiving as the pretest count suggests.

Reading a failing score report

ANCC

The handbook describes a report that lists the scaled score, the word "fail," and one of three diagnostic labels per content area:

• Low: performance below an acceptable level; substantial study recommended.
• Medium: marginally acceptable; further study recommended.
• High: well above average; a review may still help.

The labels are descriptive, not scores. A "medium" in a 50-item domain such as FNP implementation represents far more recoverable points than a "low" in a 20-item domain, so weight the labels by the size of each domain in the test content outline. The scaled score tells you distance from 350; a result in the 330s suggests a handful of items, while a result far lower suggests a content gap, not a bad day.

NPCB

NPCB candidates see a preliminary pass or fail on screen immediately, then receive an emailed score letter. The handbook states the preliminary result is not official until eligibility, exam results and documentation have been reviewed. Failing letters give feedback by domain (assess, diagnose, plan, evaluate) rather than a percentage, so the same approach applies: compare the weak domain against its item count on the blueprint to see where the points are.

What the result does not mean

• A pass is a pass. Boards of nursing receive certification status, not a score. NPCB's FAQ notes that it does not report exam failures to state boards, even if a board was selected on the application; only certification status is shared.
• Employers do not see scores. Verification of certification confirms status and dates.
• Practice-bank percentages are not predictions. A question bank's items are not equated to the live exam, and its percentage is not a scaled score. Trend lines and domain breakdowns are useful; a single number is not. Platforms such as NPPace report by blueprint domain for this reason.
• A near miss is not evidence of an error. NPCB offers a paid rescore that verifies the calculation only; ANCC's appeals policy excludes challenges to the passing score, the items or the test's validity. Scoring errors on computer-delivered exams are rare, and a rescore does not re-key items.

Preparing with the scoring model in mind

• Study to the blueprint weights, because every scored item carries equal credit and the domains differ in size.
• Treat time as a scoring resource: an unanswered item is scored as wrong, so a candidate with ten items left at the five-minute mark should answer all ten.
• Do not change answers on a hunch. Change only when a re-read reveals a missed qualifier such as "initial," "most likely" or "contraindicated."
• Track practice performance by domain and by age group. A flat 75% overall can hide a 55% in evaluation, which is where borderline results are decided.
• After a fail, request nothing until the official report arrives; the domain feedback is the retake plan.

Key points

Related NPPace guides

• AANP vs ANCC: Which NP Certification Exam Should You Take?
• AGPCNP Exam Blueprint Explained: Domains, Question Counts and Focus
• Failed the NP Exam? Retake Rules for AANP and ANCC and a 6-Week Recovery Plan
• The 8-Week NP Board Study Plan (FNP and AGNP)