The NCLEX is not graded by the percentage of questions you get right.
Try it now
In 30 seconds
- 1It estimates your ability from every scored answer and the difficulty of each item, then compares that estimate with the passing standard.
- 2After the minimum, the whole 95% range above the line is a pass, the whole range below is a fail, and a range that crosses needs more items.
- 3Near the end, candidates commonly see items they have about a 50% chance of answering completely correctly, so raw percent can mislead.
It estimates your nursing ability from every scored answer and the difficulty of the item, then compares that estimate with the passing standard. That is why two candidates can answer roughly the same percentage correctly and receive different results.
1. You answer
Right, wrong, or partial credit on an NGN item.
2. Ability updates
Every prior answer and its difficulty affect the estimate.
3. The next item is chosen
Difficulty is targeted while the content blueprint stays balanced.
4. The computer checks
Is the full confidence range above or below the passing line?
The plain-English version
Imagine a line. The left side is below the level required for safe entry-level practice; the right side is above it. After every answer, the computer places its best estimate of your ability on that line. Because an estimate is never perfectly certain, it also draws a range around the estimate.
NCLEX scoring · the 95% rule
The whole range matters, not just the dot
Dot = ability estimate. Bar = 95% estimated ability range. Illustration only.
The dot is the computer’s best estimate of your ability. The bar around it is the 95% confidence range. The vertical line is the passing standard.
After the minimum number of items, the most common stopping rule is simple: if the entire range is above the passing standard, you pass.
If the entire range is below it, you fail.
If the range still crosses the line, the computer needs more evidence, and the exam keeps going.
Make it stick. A free account turns what you just read into speed-learning rounds, practice questions and a tutor that explains.
Create my free accountWhy 50% correct can still be doing well
A CAT tries to serve items near your current ability. A strong candidate does not receive a comfortable parade of easy questions; the computer keeps moving the challenge upward.
Why raw percent can mislead
Two 50% scores can carry different evidence
Student A
50%
Half correct while the item difficulty stayed mostly below the passing standard.
Student B
50%
Half correct while the CAT kept serving items around or above the passing standard.
Illustration only. The NCLEX uses calibrated item parameters and a running ability estimate—not this bar chart or a simple difficulty average.
This is also why the last item, the number of SATA questions, and how awful the exam felt cannot tell you whether you passed. Difficulty is calibrated statistically, not by how a question feels to one person; see what shutting off at 85 does and does not mean.
The three ways the real NCLEX reaches a decision
- 95% confidence rule. After at least 85 items, the exam stops when it is 95% certain your ability is clearly above or below the standard.
- Maximum-length rule. If your range keeps crossing the line through item 150, the final ability estimate decides: at or above the standard passes; below it fails.
- Run-out-of-time rule. If time expires before the maximum, the official alternate rule applies. Fewer than the required minimum items is a fail; otherwise the final estimate from completed items is compared with the passing standard.
Why the length lands anywhere from 85 to 150 is its own post: how many questions is the NCLEX.
85
items before the 95% confidence rule can stop the exam
150
items at most; then the final ability estimate decides
95%
certainty the exam needs that you are clearly above or below the standard
Difficulty is not the only thing the CAT controls
The real exam also has to satisfy the NCLEX test-plan percentages. The official candidate bulletin describes item selection as a combination of content balance, information about ability, and exposure control for repeat candidates. Clinical-judgment coverage is built in as well: three case-study sets plus stand-alone clinical-judgment items are selected according to exam length. It does not simply chase your weakest specialty after one miss.
Quick check · no scrolling back
After 85 items, the whole 95% range sits above the passing line. What happens?
How close is the NurseSavvy CAT simulation?
NurseSavvy now uses a high-fidelity simulation of the publicly described CAT process. A new readiness exam updates an ability estimate after every scored response, uses the item’s authored difficulty and NGN partial credit, targets the next difficulty, preserves the content blueprint, and reports a 95% ability range against a passing line. The default exam is 85 items. If you enable the extension option, an uncertain test can continue toward 150. The walkthrough is in how the NurseSavvy CAT practice test works.
| Feature | Real NCLEX | NurseSavvy simulation |
|---|---|---|
| Updates after each answer | Yes | Yes |
| Targets item difficulty to ability | Yes | Yes, across 6 observed tiers |
| Balances the test-plan content | Yes | Yes |
| Three clinical-judgment case studies | Yes | Yes |
| Minimum / maximum length | 85 / 150 | 85 / 150 when extension is enabled |
| 95% confidence stopping | Yes | Yes when extension is enabled |
| Unscored pretest items | Yes | No—every NurseSavvy item counts |
| Exact items and calibration | Proprietary NCSBN pool | NurseSavvy-authored pool and calibration |
The simulation also guarantees the kinds of work a serious readiness exam needs: three six-item case studies, dosage calculations, SATA, bowties, and a paced mix of stand-alone NGN formats. NurseSavvy’s six authored difficulty tiers have shown the expected pattern in real student responses: accuracy falls as authored difficulty rises (the data is in difficulty calibration). That gives the adaptive engine a meaningful ladder to climb.
What NurseSavvy cannot—and does not—claim
No outside prep company has the exact NCLEX algorithm, operational item pool, pretest-item calibration, exposure controls, or psychometric equating. Those belong to NCSBN. A prep company can reproduce the public structure and measurement logic closely; it cannot truthfully call its engine the real NCLEX engine.
NurseSavvy is therefore a realistic CAT simulation, not an official score report or a guarantee of passing. Its result is strongest when you use the CAT confidence signal, difficulty breakdown, topic results, and repeated performance together.
How to read your NurseSavvy result
- Confidence range above the line: the test produced a likely-pass signal.
- Range crossing the line: the result is too close to call confidently; more evidence or more preparation is useful.
- Range below the line: the test produced a meaningful risk signal and your study plan should lead the next step.
- Raw accuracy lower than expected: check the difficulty mix before panicking. A CAT that adapted upward is supposed to keep the work hard.
Your next useful comparison is percentile: not “what percentage did I get right?” but “where did this performance rank among comparable test takers?” Read the visual guide to NCLEX practice-test percentiles.
Try the CAT instead of guessing from a raw score
A full NurseSavvy readiness exam is free and unlimited. It defaults to 85 items and can continue when the estimate remains uncertain.
Official sources
Common questions
What percentage do you need to pass the NCLEX?
There is no fixed percentage-correct cutoff. The NCLEX estimates candidate ability from the difficulty and scoring of completed items, then compares that estimate with the passing standard.
Can you pass the NCLEX with 50% correct?
Raw accuracy alone cannot answer that. A CAT targets questions near the candidate’s current ability, so a candidate performing above the passing standard may still find items difficult and answer only around half of targeted items completely correctly.
What is the NCLEX 95% confidence rule?
After the minimum number of items, the exam can stop when it is 95% certain that the candidate’s ability is clearly above or clearly below the passing standard.
Does NurseSavvy use computerized adaptive testing?
Yes. New NurseSavvy readiness exams use a CAT simulation that updates ability after each scored response, targets difficulty, balances the NCLEX content blueprint, and can apply a 95% confidence stopping rule when extension is enabled. It is not the proprietary NCSBN algorithm or item pool.