Nursing care
Pass-Fail Rules, explained for the bedside and the exam
Written and reviewed by Dana Whitfield, RN, MSN · 4 min read · Updated September 2026
Short answer
The exam ends one of two ways: the computer reaches 95% confidence in a pass or fail decision at a minimum of 85 questions, or it runs every candidate to 150 questions and the final items decide the result. Stopping at 85 is not automatically a fail, and running to 150 is not automatically a fail either. Both are ordinary outcomes of the same scoring rule.
The idea in one paragraph
The NCLEX is computer adaptive. Every answer changes the difficulty of the next question, and the software is constantly estimating whether your ability sits above or below the passing standard. Once that estimate is precise enough to say so with 95% confidence, the exam stops. That can happen as early as question 85.
If confidence never reaches that threshold, the exam keeps going until the maximum length, 150 questions, and the final answers tip the estimate one way or the other. Length alone tells you nothing about outcome. A candidate who stops at 85 and a candidate who runs to 150 can both pass, and both can both fail.
Why it matters clinically
This isn't an abstract testing quirk. It maps onto how the exam samples your practice across the whole NCLEX test plan, safe and effective care environment, health promotion, psychosocial integrity, physiological integrity. A short exam means the algorithm found a clear signal fast, either strongly above or strongly below the standard. A long exam means your performance has hovered close to the passing line, and the system needs more data points to resolve it.
Understanding this stops candidates from making clinical-reasoning errors about their own performance mid-exam, the same kind of premature pattern-matching you'd flag in a student jumping to a diagnosis from one vital sign. Reading too much into question count during the test wastes cognitive load you need for the next item.
How to apply it at the bedside
At the bedside, the transferable skill is the same one the algorithm is testing: treat every new data point on its own merits rather than as confirmation of a story you've already built. A nurse who decides a patient is stable after one reassuring set of vitals and stops reassessing is making the same error as a candidate who decides they're failing because the exam hasn't stopped by question 100.
Apply the discipline directly during the test. Answer each question as a discrete clinical decision, then let it go. Do not try to count questions, guess your ability estimate, or infer scoring pressure from how a question feels. None of that information is available to you in real time, and acting on a guess is worse than acting on none.
Where students get it wrong
The most common error is treating a stop at 85 as proof of failure. Candidates walk out convinced they failed because the exam felt short, and some genuinely did well enough that the system was confident early. The reverse mistake is equally common: assuming a full 150-question run means failure because the exam kept testing you. Neither inference is supported by how the algorithm works.
A second error is changing test-taking behavior based on a guess about length partway through, slowing down, rushing, or second-guessing answers because the exam hasn't stopped. That behavior change can genuinely hurt performance even though the original guess about pass or fail was wrong.
Worked examples
Candidate A answers steadily, and the exam stops at question 91. The algorithm reached 95% confidence that their ability estimate sits above the passing standard. Result: pass. Candidate B also stops at 91, but the confidence interval closed on the other side of the standard. Result: fail. From the outside, both experiences looked identical.
Candidate C runs the full 150 questions. Their performance tracked close to the passing standard throughout, so the algorithm never reached the 95% threshold early and used every available item. The final questions determined whether their ability estimate landed just above or just below the line. Length told them nothing until the result was released.
How the exam tests it
The exam does not test this rule directly, no item asks you to define CAT scoring, but understanding it changes how you should behave under exam conditions. Test-taking strategy content in NCLEX prep addresses this because self-monitoring during a long adaptive exam is itself a form of test performance.
Where it matters most is candidate behavior in the final stretch of a long exam. Fatigue and anxiety peak near question 130 to 150, exactly the range where candidates are most likely to misread exam length as a bad sign and change strategy. Prep material that flags this is preparing you for exam-day psychology, not content.
The next step on this is the same as on everything else here: answer questions and read the rationales. Our sata questions practice questions are the closest set to what this page covers.
One question from the sata questions set
A client is admitted with diabetic ketoacidosis. Which findings does the nurse expect? Select all that apply.
Rationale
DKA is hyperglycemia plus ketosis plus metabolic acidosis, and four of these are that picture: Kussmaul respirations blowing off CO₂, ketones on the breath, a glucose well above 250 mg/dL, and dehydration showing as warm, flushed, dry skin. A bicarbonate of 30 mEq/L is above the reference range — in DKA bicarbonate is consumed and falls below 18. On a select-all, check each option against the pathophysiology on its own; there is no partial credit on the real exam.
Answer: A, B, D, E
Common questions
Does stopping at 85 questions mean I failed?
No. Stopping at the minimum of 85 means the algorithm reached 95% confidence in a result, and that result can be pass or fail. Length alone never tells you which.
Is running to 150 questions a bad sign?
Not on its own. It means your ability estimate stayed close to the passing standard for longer, so the exam needed more items to resolve a confident result. Many candidates who run the full length pass.
Can I tell how I'm doing from how hard the questions feel?
No. Question difficulty adapts continuously and isn't a reliable signal of your standing. Treat each question independently and avoid drawing conclusions mid-exam.
What's the actual maximum number of questions?
150 questions is the maximum for the NCLEX-RN under current CAT rules; the minimum is 85. Exact minimums and maximums are set by the testing body and can be updated, so confirm current figures before exam day.
More on sata questions
Guides on this