What is the passing score for certification exams?
Passing scores vary by exam and are set through a standard-setting process by the certifying body. Most are not simple percentages -- many use scaled scores. The AWS SAA-C03 requires a scaled score of 720 out of 1000; CompTIA exams require 750 out of 900; the CISSP uses a scaled pass/fail with a cut score around 700 out of 1000. Check your specific exam's candidate handbook for the official passing threshold.
Understanding how certification exam scoring works removes one of the most common sources of candidate anxiety: uncertainty about what is actually required to pass. Many candidates study toward a mental target of "80%" or "90%" without understanding that these numbers have no direct relationship to how their exam is actually scored.
This guide explains scaled scoring, standard-setting methodology, adaptive exam scoring, and what your practice exam percentage actually predicts about your real exam performance.
Scaled Scores vs. Raw Scores
Most professional certification exams do not report or use raw percentage scores. They report scaled scores -- numbers on a proprietary scale that allow comparison across different versions of the same exam.
Why use scaled scores? Because certification exams are regularly updated. Question pools rotate; some versions of an exam are slightly harder than others. Scaled scoring adjusts for these differences so that a passing score of 720 means the same thing regardless of which question version you received.
| Exam | Score Scale | Passing Score |
|---|---|---|
| AWS Solutions Architect Associate | 100-1000 | 720 |
| CompTIA Security+ | 100-900 | 750 |
| CompTIA Network+ | 100-900 | 720 |
| CISSP | 250-1000 | 700 |
| PMP | Pass/Fail with domain percentiles | Pass threshold not published |
| Microsoft AZ-900 | 0-1000 | 700 |
| Google Cloud Associate | 0-200 (approx.) | Not published |
| CCNA | Not published | Not published (around 825/1000 historically) |
The practical implication: when you score 750 on a CompTIA exam, this does not mean you answered 83% of questions correctly. The scaled score is derived from item response theory calculations that weight questions by difficulty.
How the Passing Score Is Set
Passing scores are established through a standard-setting study conducted by the certifying body before the exam is released. A panel of subject matter experts (SMEs) evaluates each exam question and determines what percentage of minimally competent candidates (people who just barely deserve to pass) would answer correctly.
This process, called the Angoff method in most certifying bodies, produces a criterion-referenced passing score. It is not norm-referenced (based on how test-takers perform relative to each other) -- it is criterion-referenced (based on what a minimally competent professional should know).
The implications:
- You are not competing against other candidates. The passing score does not change based on how many people fail. You either demonstrate minimum competency or you do not.
- The passing score is not arbitrary. It reflects a deliberate judgment by domain experts about what entry-level certification holders should know.
- Slightly harder exam versions produce the same scaled pass threshold, because the scaling process adjusts for difficulty.
"Standard-setting for professional certifications is a rigorous psychometric process. The Angoff and related methods produce defensible minimum competency thresholds that are criterion-referenced to job performance requirements. Candidates who meet the threshold have demonstrated the knowledge and judgment required for the role the certification represents." -- Dr. Gregory Cizek, University of North Carolina Chapel Hill, Standard Setting: A Guide to Establishing and Evaluating Performance Standards on Tests
Adaptive Exam Scoring: How CISSP and Similar Exams Work
Computer Adaptive Testing (CAT) exams work differently from linear exams. The CISSP uses a CAT format where the exam adapts difficulty based on your answers.
Key properties of CAT exams:
- The exam starts at medium difficulty
- Correct answers lead to harder questions; incorrect answers lead to easier questions
- The exam continues until it is statistically confident you are above or below the passing threshold
- The CISSP can end as early as 100 questions (out of a possible 175) if the system reaches confidence
What this means practically:
- You cannot estimate your score from the number of questions you received. Ending at 100 questions does not indicate pass or fail -- it means the system reached statistical confidence in either direction.
- Hard questions are a positive signal. If your exam is delivering hard questions, you are likely performing above the threshold. CAT systems increase difficulty when you answer correctly.
- You cannot tell mid-exam if you are passing. The adaptive algorithm is opaque; trying to infer your standing from question difficulty is not reliable.
"In computer adaptive testing, the exam's purpose is not to measure what you know -- it is to determine whether you meet the minimum competency standard as efficiently as possible. The algorithm terminates when statistical confidence is achieved. Candidates who exit early (at minimum question count) have not been 'screened out' -- they have simply given the system enough signal to reach a decision." -- ISACA CAT Implementation Documentation
What Your Practice Exam Score Actually Means
Practice exam percentage and scaled exam score are related but not directly comparable. The relationship depends on:
- Practice exam quality: High-quality providers whose questions match real exam difficulty produce more predictive scores. Generic practice questions do not.
- Question pool alignment: Practice exams that align to the current exam objectives and question style are better predictors than outdated question banks.
- Testing conditions: Practice scores under timed, realistic conditions are more predictive than untimed, open-book review.
General guideline for predicting real exam performance:
| Practice Exam Score (on-target provider) | Real Exam Outcome Prediction |
|---|---|
| Below 65% | Not ready; significant gaps remain |
| 65-74% | Borderline; some domains need reinforcement |
| 75-84% | Likely ready; verify weak domains |
| 85%+ | Well-prepared; maintain and avoid burnout |
These are directional, not guarantees. The variance is significant -- some candidates who practice at 75% pass comfortably; some who practice at 85% fail on exam day due to test anxiety, unfamiliar question phrasing, or encountering heavy weight in a weak domain.
Score Reports After the Exam
For exams that do not announce pass/fail immediately (some proctored exams), score reports arrive within hours to weeks depending on the provider. Most Pearson VUE exams show pass/fail on-screen immediately after submission.
Score reports typically include:
- Overall pass/fail status
- Scaled score (where applicable)
- Domain or section percentile performance
Using domain scores for retake planning: If you fail, the domain breakdown tells you where to focus. A score report showing 60% in one domain and 90% in others tells you exactly where your second attempt study effort should go.
Frequently Asked Questions
If I answer 80% of questions correctly, will I pass? Not necessarily. With scaled scoring, the relationship between raw percentage correct and scaled score depends on question difficulty weighting. A 80% on difficult questions may produce a higher scaled score than 80% on easy questions. Focus on the scaled passing threshold, not a target percentage.
Does the passing score change based on how hard the exam is? The scaled passing threshold stays the same, but the raw percentage required to reach it varies with question difficulty. On a harder exam version, you may need a lower raw percentage to achieve the same scaled score. This is the purpose of scaling.
What happens when I fail by a small margin? Most certifying bodies allow retakes after a waiting period (typically 14-30 days for a first failure, longer for subsequent failures). The waiting period is not punitive -- research on exam performance shows that attempting a retake without additional study produces minimal improvement. Use the score report's domain breakdown to guide targeted study before your retake.
References
- Cizek, G.J., & Bunch, M.B. (2007). Standard setting: A guide to establishing and evaluating performance standards on tests. SAGE Publications.
- Angoff, W.H. (1971). Scales, norms, and equivalent scores. In R.L. Thorndike (Ed.), Educational Measurement (2nd ed.). American Council on Education.
- van der Linden, W.J., & Hambleton, R.K. (1997). Handbook of modern item response theory. Springer.
- ISACA. (2024). CISM/CISA computer adaptive testing candidate information guide. ISACA official documentation.
- ISC2. (2024). CISSP examination information and passing score methodology. ISC2 official documentation.
- AWS. (2024). AWS certification exam guide: Passing scores and scoring methodology. Amazon Web Services official documentation.