AWS exam passing scores explained: how the 100–1000 scale works
AWS exams are scored on a scaled range of 100 to 1000, and the passing thresholds are 700 for Cloud Practitioner, 720 for the Associate exams, and 750 for Professional level. Two properties of this system matter more than the numbers themselves: the score is scaled, meaning it is not a raw percentage of questions answered correctly, and it is compensatory, meaning only your whole-exam total counts — there is no requirement to pass each domain individually, so strength in one area can carry weakness in another. Add the fact that a portion of the questions on your paper are unscored pretest items, and the arithmetic becomes kinder than most candidates assume: you can miss a fair number of questions and still pass comfortably. This guide explains what the scale actually measures, how compensatory scoring changes study strategy, and what genuine readiness looks like before you book.
The scale: 100 to 1000, and what the thresholds are
Every AWS exam reports a single scaled score between 100 and 1000, with the pass mark depending on the level: 700 for the foundational Cloud Practitioner exam, 720 for the Associate exams — Solutions Architect, Developer, and CloudOps Engineer — and 750 for Professional level. You receive the number itself with your result, so you always know your margin, whether comfortable or narrow.
One thing the number is not: a ranking against other candidates. AWS exams are not graded on a curve, and there is no quota of passes per sitting — every candidate who reaches the threshold passes, regardless of how anyone else performed that day. The threshold represents a fixed standard of competence for the level, which is why it differs by tier: the bar rises from 700 to 720 to 750 as the exams move from foundational recall toward professional-grade judgement.
Why a scaled score is not a percentage
The most common misreading of AWS scoring is treating 720 out of 1000 as “72% of questions correct”. It is not. AWS assembles each candidate’s exam from a large pool of questions, and different papers contain different questions of different difficulty. To keep the pass standard fair across papers, raw performance is converted onto the common 100–1000 scale, with the conversion accounting for the difficulty of the specific questions you received. A slightly harder paper needs slightly fewer raw correct answers to reach the same scaled score, and vice versa.
The practical consequence is that nobody outside AWS can tell you the exact number of questions you must answer correctly — it varies by paper, and AWS does not publish the conversion. This is liberating rather than frustrating once accepted: chasing a precise raw-mark target is impossible, so the only workable strategy is building a comfortable margin. Aim to perform clearly above the threshold in practice, and the scaling arithmetic on the day takes care of itself.
Compensatory scoring: the whole exam counts, not each domain
AWS scoring is compensatory: your pass or fail is decided by the whole-exam scaled score alone, and there is no requirement to clear a minimum in every domain. A strong performance in a heavily weighted domain can offset a genuinely weak one elsewhere, and candidates pass every day with one domain they would privately admit was shaky.
This should shape how you study, in both directions. It means perfectionism across every domain is unnecessary — you do not need to master the smallest domain to its corners before booking. But it also rewards prioritising by weight: the exam guide states each domain’s share of the exam, and marks in a domain worth a third of the paper are worth proportionally more study time than marks in one worth a tenth. The rational plan is solid coverage everywhere, depth where the weight is, and acceptance that one modest domain is survivable if the heavyweights are strong.
The score report, and what readiness looks like beforehand
Alongside the scaled score, your result includes domain-level performance feedback — an indication of which domains were strong and which need improvement, rather than exact per-domain marks. After a pass it is a map of what to shore up before the knowledge matters at work; after a fail it is the study plan for the retake, telling you precisely where the marks went missing. Either way, read it — the single number decides the outcome, but the breakdown is the useful part.
Before the exam, the honest question is how to know you would clear the threshold, and the answer is measurement, not feeling. Full-length, timed mock exams are the closest available estimate of exam-day performance, and the readiness signal is consistency: clearing complete mocks repeatedly, at a margin comfortably above the pass mark, including in your weaker domains. One good mock can be luck; a run of them is evidence. When your practice results stop being an argument and start being a pattern, the scaled score on the day tends to follow — and if the attempt does go wrong, our guide on what happens if you fail an AWS exam covers the recovery route.
Original practice questions, timed mock exams and revision notes. No card, nothing to pay.