The NMC publishes no raw pass mark on its candidate pages. The 2021 test specification says the standard is set by the Angoff method and maintained statistically, with a Part A pass mark of 87% (13 of 15) and a Part B pass mark generally 60% to 70%, depending on field. Each part is passed or failed separately and both are needed. Quarterly pass rates in 2025 ran from 76% down to 65% for candidates passing both parts, and 63% of 2,967 applicants passed in January to March 2026.
What the NMC actually states
Search for the NMC CBT pass mark and you will find confident numbers on forums and prep sites. The NMC candidate pages do not give a raw percentage. The two official documents that mention a standard are the Test of Competence 2021 test specification for nursing and the CBT information booklet. The booklet confirms that you pass or fail each part separately. The specification explains how the standard is set.
The specification says the pass mark is set with the Angoff method and then maintained statistically from one test version to the next. It gives two reference points: 87% for Part A, which is 13 of 15 questions, and generally 60% to 70% for Part B, depending on the field of practice. We checked these against the document on 25 October 2026. The document is dated 2021, and we found no later publication that re-confirms the figures, so use them as planning targets, not as a promise.
Part A at 13 of 15 is 86.7% in arithmetic terms. The 87% in the specification is a rounded figure for the same standard. Two wrong answers is the maximum you can afford.
How the Angoff method works
In an Angoff standard-setting exercise, a panel of expert judges looks at each question and estimates the probability that a borderline candidate, someone just competent enough to pass, would answer it correctly. The estimates are added across the test to give the score a borderline candidate would be expected to reach.
That has three practical consequences for you.
- The standard is about competence, not ranking. A pass is not a quota. You are not competing against the other people in the room.
- Harder papers have lower raw cut scores. If judges rate a version of the test as harder, the raw score needed goes down. The standard of performance stays the same. This is why the "number you need" can differ between candidates who sat on different days.
- Judges differ by item. Questions on well-known procedures have high borderline probabilities; subtle law or evaluation questions have lower ones. A topic that feels simple to you may still carry weight.
Because the NMC maintains the standard statistically, you cannot calculate your own pass mark from the question count. A claim such as "you need exactly 62 of 100 correct" is a guess.
Two parts, two decisions
The CBT combines two tests in one sitting of 3 hours: Part A numeracy with 15 questions in 30 minutes, and Part B clinical with 100 questions in 2 hours 30 minutes. Each is scored against its own standard. Doing well in Part B does not offset a Part A failure, and the reverse is also true.
| Item | Part A | Part B |
|---|---|---|
| Questions | 15 | 100 |
| Standard in 2021 specification | 87% (13 of 15) | Generally 60% to 70%, varies by field |
| Margin for error | 2 wrong answers | 30 to 40 wrong answers, but unknown for your version |
| Answer format in the real exam | Typed numeric answer | Four options, one most correct |
| Resit fee (checked 25 Oct 2026) | £50 | £70 |
The Part B margin row is illustrative. It shows the arithmetic of the range quoted in the specification and not a promise about your paper. Treat the range as a ceiling for planning: if you consistently score above 75% on realistic mocks, you have a buffer, and below 65% you do not.
Both parts must be passed for the CBT to count. Your NMC Online account shows Pass only when both are passed, and shows the overall result only. For the part-by-part breakdown, use the Pearson VUE account, which updates within 48 hours.
What the pass rates say
The NMC publishes quarterly pass rates for the CBT. They include all fields of practice, so adult nursing candidates should not read them as exact. The numbers are still the best public evidence of difficulty.
| Period | Part A first attempt | Part B first attempt | Passed both parts |
|---|---|---|---|
| Jan to Mar 2025 | 89% | 84% | 76% |
| Apr to Jun 2025 | 88% | 81% | 74% |
| Jul to Sep 2025 | 87% | 80% | 71% |
| Oct to Dec 2025 | 76% | 77% | 65% |
| Jan to Mar 2026 | 77% | not shown on the page we read | 63% of 2,967 applicants |
Four points follow from the table.
- First-attempt Part A fell in late 2025. It moved from 87% to 89% in the first three quarters of 2025 down to 76% in October to December, and stayed at 77% in early 2026. The NMC does not explain the fall, and neither do we.
- Resits pass at lower rates. In January to March 2026 Part A passed at 77% on a first attempt, 63% on a second and 59% on a third. For Part B in October to December 2025, the second attempt passed at 64% and the third at 54%.
- Overall figures include repeat candidates. The share of all applicants passing both parts is lower than the first-attempt share because retakers dilute it. In early 2026, first-attempt candidates passed at 76% overall from 2,150 candidates.
- Numbers move quarter to quarter. Do not extrapolate. Use them to judge seriousness, not to predict your own result.
The scoring details we cannot confirm
Several questions come up repeatedly in candidate forums. The NMC pages we read do not answer them, so we do not either.
- Whether wrong answers lose marks. Not stated. Answer every question.
- Whether the exam includes unscored trial items. Not stated on the pages we read.
- Whether the Part B standard differs for adult nursing within the 60% to 70% range. The specification says it varies by field and does not give field figures.
- Whether Part A partial credit applies. The real Part A answer is a single typed number, and the NMC gives no partial credit rules. Assume a wrong digit is a wrong answer.
- Your raw score. The NMC says no further feedback is given beyond the part breakdown.
Sit a timed NMC CBT mock exam
Blueprint-weighted questions under exam timing, then a breakdown of your result by domain.
How to turn the standard into a target
You cannot see the standard, so you build a safety margin. The margin is cheaper to build in Part A than in Part B.
Part A target
Aim for 15 of 15 on untimed practice and 14 or more on timed practice. A candidate who averages 13 on timed practice is at the pass line with no room for a typing slip. The question types are narrow (dose, units, oral medicines, injections, infusions, fluid balance), which is why this part responds well to deliberate practice. See how to pass Part A numeracy.
Part B target
For clinical practice, use a score of about 75% on mixed, timed sets as your readiness line, with no domain below 65%. That figure is our editorial judgement, set above the upper end of the quoted range to allow for the difference between a practice bank and a live paper. Check it against your error log: if most misses are guesses between two plausible options, you are close; if they are knowledge gaps, you are not.
Domain floors
The blueprint gives Platform 1 17 questions and Platform 4 15. A weak domain at the top of the table costs more than a weak domain at the bottom. See the hub blueprint for the full weights.
What forum claims often get wrong
- "You need 60% in Part B." The specification gives a general range of 60% to 70% that varies by field, not a single figure.
- "Numeracy is 80%." The 2021 specification states 87%, which is 13 of 15.
- "Total score decides." The two parts are separate. A high Part B does not rescue Part A.
- "Pass rates are fixed." The NMC quarterly figures moved from 76% to 65% across 2025, which fits a competence standard and not a quota.
What a borderline result tells you
Your Pearson VUE report splits the result by part. If you failed Part A only, the plan is numeracy, and you pay the Part A resit fee. If you failed Part B only, the plan is domain repair. If you failed both, plan both, and note that resitting both parts costs the same as the first booking. Rules on attempts, spacing and costs are in resits and attempts.
Do not buy a plan based on a claimed pass mark. If a seller states a precise percentage for the NMC CBT, ask for the NMC source. We found none beyond the 2021 test specification.
Pass marks matter less than process. The candidates who pass are the ones who fix their specific errors early. Our study plan builds the error log into every week.
Frequently asked questions
What is the NMC CBT pass mark for numeracy?
The 2021 test specification gives a Part A pass mark of 87%, which is 13 of 15 questions. We found no 2026 document re-confirming it, so treat it as the planning target and aim for 14 or 15.
What is the Part B pass mark?
The 2021 specification says generally 60% to 70%, depending on the field of practice, with the standard set by the Angoff method and maintained statistically. The NMC does not give a figure for adult nursing on its candidate pages.
Do I need to pass both parts?
Yes. Part A and Part B are passed or failed separately and both must be passed for the CBT to count. Your NMC Online account shows Pass only once both parts are passed.
Do wrong answers lose marks?
The NMC pages we read do not say. Answer every question, because a blank cannot score.
What was the NMC CBT pass rate in 2025?
Candidates passing both parts: 76% in January to March, 74% in April to June, 71% in July to September and 65% in October to December 2025. Figures cover all fields and all attempts.
Is the Part A pass rate lower on resits?
Yes. In January to March 2026, Part A passed at 77% on a first attempt, 63% on a second and 59% on a third. This is why a first-attempt pass is the best use of your attempts.