AWS Cloud Practitioner Passing Score and Scaled Scoring

The passing standard is 700 on a scaled score

AWS reports CLF-C02 results on a scaled range from 100 to 1,000. A score of 700 or higher passes; a score below 700 does not. The score is not the number of questions answered correctly. There are only 50 scored questions, so 700 cannot be 700 raw answers. It is also not a raw percentage: 700 divided by 1,000 is not a statement that the candidate answered 70 percent correctly.

The distinction matters because CLF-C02 presents 65 questions, of which 50 affect the score and 15 are unscored. Candidates do not know which items are unscored. A raw count from a practice set cannot be converted into a reliable official score using simple arithmetic. AWS publishes the scaled score and pass/fail result, not a candidate-facing raw-to-scaled conversion chart.

Why AWS uses scaled scores

AWS may use multiple forms of an exam, and different forms can vary slightly in difficulty because they contain different questions. A scaled score gives results a consistent interpretation across these forms. AWS describes setting a passing standard through a modified Angoff process with content experts, then using statistical equating to determine whether the standard needs adjustment for other forms. The adjusted raw standard is converted to the reported scale score.

In practical terms, seven correct answers out of ten easy questions do not necessarily represent the same competency as seven correct out of ten harder questions. Equating accounts for differences between forms so candidates are judged against the same competency standard. AWS's public explanation gives the method at a high level; it does not provide the item-level difficulty values, candidate raw cut score, or formula needed to reverse the score.

Do not claim that every form has a known raw passing number such as 35 of 50 or 70 percent. Even if a practice resource uses a percentage target, that is its own readiness guideline. It does not reveal AWS's form-level passing standard or the candidate's scaled score.

Why 700 is not 70 percent correct

A score range beginning at 100 is not a percentage scale beginning at zero. Dividing 700 by 1,000 ignores the scale's lower bound and the conversion from raw performance to scaled performance. Even calculating (700 minus 100) divided by (1,000 minus 100) would not establish a raw percent correct, because scaled score points do not equal question-count percentage points.

Suppose one candidate earns a 700 on one form and another earns a 700 on a different form. The reported scores have the same interpretation under AWS's equating process, even if the raw number correct differed. That is the purpose of the scale. It is therefore inaccurate to say the candidate needs 70 percent of the 50 scored items or 45.5 questions. AWS has not published that conversion.

Section feedback is not a separate pass requirement

AWS's exam guide says the score report may show classifications of performance at the section level. These classifications provide general feedback about strengths and weaknesses; they are not a detailed question review or a domain-by-domain gradebook. CLF-C02 uses compensatory scoring. A candidate does not need a passing score in each domain and passes by meeting the overall scaled standard.

The four domains have different weights: Cloud Concepts (24%), Security and Compliance (30%), Cloud Technology and Services (34%), and Billing, Pricing, and Support (12%). Those weights indicate relative coverage of scored content. A weaker section may be offset by stronger performance elsewhere under the overall scoring approach, but no candidate should treat a section classification as a guarantee that another area can be ignored.

For example, a candidate might receive a passing total and see an area for improvement in Billing, Pricing, and Support. That feedback does not change the pass into a fail. It can guide continued learning in cost management or support options. Another candidate may fail overall despite strong feedback in three sections; the overall score remains decisive.

How unscored questions affect interpretation

Fifteen of the 65 presented items are unscored pretest questions. AWS uses these items to evaluate questions for possible future scored use. They are not identified during the exam. Candidates should not attempt to identify them by topic or difficulty, because the exam does not tell you which items count.

Only 50 questions affect the score, but the candidate sees and answers all 65. The public score report does not identify the unscored items or give a raw count for only the scored questions. This makes post-exam self-estimates especially unreliable: a candidate cannot know which remembered items affected the score, and memory of the session may be incomplete.

Use practice results without converting them to AWS scores

A practice set is useful for diagnosing gaps. If a candidate gets 14 of 20 security questions correct, the immediate question is which concepts caused the six misses: IAM least privilege, root account protection, service responsibilities, logging, encryption, or compliance resources. It is not whether 70 percent predicts a 700 scaled result. Review each explanation and test the same concept in new scenarios.

Practice accuracy also depends on the resource, topic mix, difficulty, repetition, and whether the candidate used notes. A percentage from a short set with repeated questions has different meaning from a fresh timed exam. Record context and use trends to judge learning, but keep the official scale interpretation separate.

1. Use a diagnostic to identify domains and task statements that need work.
2. Review explanations for wrong answers and correct guesses, not just the final percentage.
3. Complete fresh mixed practice to test whether knowledge transfers between service categories.
4. Use a timed full practice exam to develop pacing and stamina.
5. Interpret the AWS result only from the official pass/fail and scaled score report.

The score report and result timing

AWS posts final results to the Certification Account within five business days after the exam closes, except when results are under security or technical review. The score report is available in the account's Exam History for exams that use scaled scores. Candidates can view, download, or print it. The report presents the pass/fail result and scaled performance; it does not provide an answer key for each item.

If the account shows a result delay beyond the normal period, check email and the AWS Certification Account for a notice before contacting support. A review for security or technical reasons is an exception to the five-business-day timeline. A candidate should not assume a delay itself indicates a pass or fail.

What to do with a fail result

A candidate who scores below 700 should use section feedback and study records to build a new preparation plan. AWS permits a retake after 14 calendar days, with no limit on attempts, but requires the full fee for each attempt. The next plan should identify weak concepts and improve question reading, rather than simply repeating a practice test until its answers are familiar.

A candidate with a pass has met the exam standard. The score is not a complete measure of every AWS skill and does not certify architecture implementation or troubleshooting. The exam validates foundational knowledge, while work experience and further certifications can demonstrate deeper practical ability.

How AWS establishes and equates a passing standard

AWS says the passing standard for an initial test form is set using the modified Angoff method. A panel of subject-matter experts estimates how difficult each question would be for a minimally qualified candidate. After discussion and multiple rounds, the panel's ratings are used to establish a minimum raw passing standard for that form. AWS then statistically equates other forms to determine whether the raw standard needs adjustment because of differences in item difficulty.

This process is meant to preserve the same competency standard when candidates receive different forms. It does not mean one form is necessarily easier for every person, and AWS does not publish the raw cut score for a given form. The scaled result expresses performance on a common reporting scale after the form-level process. A candidate should use the published score and pass/fail result rather than speculate about the behind-the-scenes raw threshold.

Why 100 is the scale floor

The reported range begins at 100, not zero. AWS describes a scaled range and notes that it is similar to converting between measurement scales. The numerical distance between score points does not translate directly to equal numbers of correct answers. For this reason, a score of 700 is not the fraction 700/1,000 of all available points in a raw test, and 700 is not a literal count of correct responses.

Nor should a candidate assume that the lowest reported score means zero correct or that each 10 scaled points represents one additional answer. The public policy does not disclose that relationship. Score reports should be interpreted as the standardized measurement AWS provides: a score below 700 is below standard; 700 or higher meets it.

Worked example: why two raw totals might not be equivalent

Imagine two different exam forms, each with 50 scored questions. Candidate A answers 36 correctly on Form A, while Candidate B answers 36 on Form B. If Form A's items are easier on average, the same raw total might not represent the same competency as 36 on Form B. Equating considers form difficulty and converts performance so that the reported scaled scores have a consistent interpretation. This example explains the need for equating; it does not claim what either candidate's actual CLF-C02 score would be.

Now consider a practice bank where every candidate receives the same 50 easy questions. A 36-correct result is simply 72 percent on that practice set. It cannot be treated as the official score because the question difficulty, item selection, weighting, and scoring process differ. Use that result to review missed concepts, then assess readiness with fresh questions across the exam domains.

Read section classifications as broad feedback

AWS may report section-level classifications that show relative strengths and weaknesses. A classification is not a second pass/fail result, and it does not expose the count of correct answers in a domain. The weighted outline helps explain why sections contain different amounts of scored content, but the individual report is not a detailed calculation worksheet.

For example, a candidate may pass overall and receive a weaker classification in Security and Compliance. That does not undo the pass. It suggests that the candidate could benefit from additional learning in shared responsibility, IAM, encryption, AWS compliance resources, or security services. A candidate who fails can use a stronger classification in Cloud Concepts to preserve that knowledge while focusing the retake plan elsewhere.

Because the score is compensatory, there is no need to satisfy a separate domain minimum. But the weighted domains still matter in preparation. Cloud Technology and Services is 34 percent of scored content, Security and Compliance is 30 percent, Cloud Concepts is 24 percent, and Billing, Pricing, and Support is 12 percent. A candidate who ignores the lowest-weight section can still miss enough scored content to affect the overall result.

How to make practice results actionable

1. Record the source, topic mix, and whether the set was timed or open book.
2. Review all wrong answers and correct guesses, and identify the concept behind each distractor.
3. Track errors by domain and specific task rather than relying only on one aggregate percent.
4. Practice a new scenario that tests the same concept with different service names or business context.
5. Use a full timed set to test pacing, then review the official result only through AWS's score report.

Suppose you answer 18 of 25 service questions correctly. The next step is to find out whether the misses involved compute, databases, storage, networking, analytics, or service integration. If most errors confuse CloudTrail with CloudWatch, make a clear comparison and test it in a different scenario. This is concrete learning progress. Converting 18/25 into a projected 720 would be unsupported.

The useful interpretation

Read 700 as the minimum passing scaled score on AWS's 100-to-1,000 scale. It is not 700 correct answers or 70 percent raw accuracy. Scaled scoring and equating allow AWS to compare forms fairly. Domain feedback helps candidates plan study, but the overall score determines pass or fail. This interpretation keeps preparation data useful without making claims the published scoring system does not support.