No. The USMLE is not curved, and the programme says so directly: scores and pass or fail decisions are not determined using a curve. Each Step has a passing standard set in advance as a minimum level of proficiency, and everyone who reaches it passes, whether that is most examinees or nearly all of them. That single fact removes a surprising amount of anxiety, and it should change how you think about practice scores, hard forms and the people you study with. The details below are as of November 2026.
What “not curved” actually means
A curved exam ranks examinees against each other and passes a set proportion. The USMLE does the opposite. Its passing standard is a fixed threshold, and the programme states that examinees who meet it will pass regardless of the number or percentage of examinees who pass overall.
So the exam is not a competition for a limited number of passes. A cohort of unusually strong students does not raise the bar, and a weak cohort does not lower it. The only performance that affects your result is your own.
Who sets the standard, and how
The passing standards are set by the USMLE Management Committee. The committee does not pick a number in isolation. Its reviews draw on recommendations from panels of physicians and educators, surveys of people who use the scores, such as licensing authorities and residency programme directors, trends in examinee performance and the precision of the scores themselves.
Each Step’s standard is reviewed periodically, typically every few years, and can be adjusted. When a review is scheduled or a change is made, the USMLE announces it on its website along with the date from which any new standard applies. As of November 2026, for example, the programme has announced a scheduled review of the Step 3 passing standard, with any approved change applying to examinees who test on or after 1 January 2027.
Why a harder form does not hurt you
Not everyone sits the same set of questions. The USMLE uses multiple forms, and some will inevitably be slightly harder than others. Without a correction, that would be unfair.
The correction is called equating. It adjusts scores for differences in question difficulty between forms, so a given score represents the same level of performance whichever form you sat. In practical terms, the percentage of questions you need to answer correctly to pass varies from form to form, but the level of proficiency it represents does not.
This is the answer to a worry that comes up after almost every exam: “my form felt much harder than my friend’s”. It may have been. Equating is there precisely so that it does not matter.
Unscored questions are not a curve either
Each exam also includes pretest questions that do not count towards your score. They are new items being checked before future use. They sometimes get confused with curving, but they are unrelated: they simply do not contribute to your result, and you cannot tell which ones they are.
What this means for practice exams
Because the passing standard is fixed, the right question about your preparation is not “how am I doing compared with everyone else?” It is “am I reliably above the standard, with margin?”
That is exactly what good practice assessments tell you. NBME self-assessments report results in relation to the real exam, and a trend of results comfortably above the passing level across several assessments is the evidence that matters. A single result, or a comparison with a classmate’s result, is not.
It also means that the percentage correct in your question bank is a weak guide on its own. Bank questions are written at varying difficulty, and your percentage depends as much on which questions you happened to do as on what you know. Banks that report performance by system and discipline, as UWorld, AMBOSS and QBankly all do in their own ways, are more useful for finding weak areas than for predicting a pass.
What this means for study groups
A curve would make classmates rivals. A fixed standard makes them allies. Sharing resources, explaining topics to each other and comparing notes on weak areas costs you nothing, because someone else’s pass does not take a pass away from you.
The one comparison worth avoiding is between scores. It tells you nothing about your own distance from the standard, and it reliably makes people feel worse.
The practical takeaway
The USMLE asks one question of each examinee: did you reach the standard? Prepare to answer it. Track your own practice assessment trend against the passing level, fix the weaknesses it shows, and stop worrying about forms, cohorts and how everyone else is doing. None of those move the line.
Frequently asked questions
Is Step 1 graded on a curve?
No. Step 1 is reported as pass or fail, and the pass is decided by a fixed minimum standard rather than by comparison with other examinees. If you meet the standard you pass, however many other people also pass.
Does it matter who else takes the USMLE on the same day?
No. Your result depends on your own performance against the passing standard. Other examinees' results do not move the line, and equating corrects for differences in difficulty between the forms people sit.
How often does the USMLE change the passing score?
The USMLE Management Committee reviews each Step's passing standard periodically, typically every few years. Reviews and any resulting changes are announced on the USMLE website, with a date from which the new standard applies.