You finish your shelf exam and see a number on your report. Maybe it’s a 72. Maybe it’s a 78 or an 84. The first questions most students ask are simple. Is that good? What percentile is it? Is it enough for Honors? And does a 78 mean the same thing on Internal Medicine as it does on Psychiatry?
Here is the short answer. An NBME shelf score is not just the percentage of questions you got right. The same number also does not mean the same thing on every subject.
NBME reports shelf performance using an Equated Percent Correct (EPC) score. This score is adjusted for small differences in difficulty between exam forms. Your percentile then depends on the subject, the comparison group, and sometimes the time of year you took the exam.
This guide will walk you through EPC scores and speciality-specific percentile benchmarks, whether shelf scores matter for Step 2 CK, and what international medical graduates should know. So, read till the end.
An NBME shelf exam is the common name for an NBME Clinical Science Subject Exam. Medical schools use these exams to test what students have learned during a clinical clerkship.
Most students take a shelf exam at the end of each core rotation, usually during the third year of medical school. The exam is written by NBME, but your school decides how to use the result. In most schools, the shelf is one part of your final clerkship grade, along with clinical evaluations and other assessments.
The name “shelf” comes from the idea that these are ready-made, “off-the-shelf” exams that any participating school can use.
NBME offers clinical science subject exams in the core clerkship areas. These are the ones students talk about most:
| Subject | Common name |
| Internal Medicine | IM shelf / Medicine shelf |
| Surgery | Surgery shelf |
| Pediatrics | Peds shelf |
| Psychiatry | Psych shelf |
| Obstetrics & Gynecology | OB/GYN shelf |
| Family Medicine | FM shelf |
| Neurology | Neuro shelf |
| Ambulatory Care | Ambulatory Medicine shelf |
Not every school uses every exam. Your clerkship director will tell you which shelf exams are part of your curriculum.
This is the most important part of understanding your score. It helps to think of scoring as a chain of steps:
Questions answered → raw performance → Equated Percent Correct (EPC) → percentile rank → your school’s clerkship grade
NBME controls the first few steps. Your medical school controls the last one.
According to NBME’s Clinical Science Subject Exam Score Interpretation Guide, the EPC score represents the percentage of the content you have mastered. It is reported on a familiar 0 to 100 scale.
The key word is “equated.” NBME uses several different versions, or forms, of each shelf exam. One form may be slightly harder than another. Equating is a statistical adjustment that corrects for these small differences. The goal is simple: two students with the same level of knowledge should get about the same EPC score, even if they took different forms.
No. They are close, but they are not the same thing.
Here is a simple example. Say you answered 80% of the scored questions correctly. Your reported EPC may not be exactly 80. If your form was a little harder than average, your EPC might be slightly higher. If your form was a little easier, it might be slightly lower.
There are two more reasons your raw percentage and your EPC may not match:
So the number of questions you saw on test day may be larger than the number that were actually scored. This is one reason you can’t calculate your EPC by simply counting what you think you got right.
A percentile rank tells you how your score compares with other students.
If you are at the 75th percentile, your score was at or above the scores of about 75% of the students in the comparison group. It does not mean you answered 75% of the questions correctly.
For shelf exams, NBME builds these comparisons, called norms, from first-time takers who took the exam at the end of a clerkship. NBME provides norms for the full academic year and also by quarter.
NBME adjusts for difficulty through equating, not through grading on a curve. Equating makes scores fair across different exam forms. It does not force a certain number of students into each score range.
NBME also reminds schools that all scores have some imprecision. For example, the Psychiatry guide lists a standard error of about 4 points. In plain terms, if you took the exam again without learning or forgetting anything, your score would usually land within about 4 points of your first result. That is worth remembering before you worry about a one- or two-point difference.
This is one of the most searched questions about shelf exams, and the honest answer is: not in the way most students mean it.
When students say “curved,” they usually mean that scores are pushed up or down so a fixed share of the class gets each grade. That is not how NBME scores the shelf. Here is what actually happens:
So there are really two different ideas here. Your score is adjusted for exam difficulty. Your percentile is relative to other students. Neither one is a classic class curve. Many articles online blur these ideas together, which is why students often get confused.
These terms often get mixed up. Let’s separate them clearly.
A raw score is your direct performance: how many scored questions you answered correctly. Students usually don’t see a true raw score on their shelf report, and it isn’t the number schools use for grading.
The EPC is the main number on your shelf report. It estimates the percentage of the subject’s content you have mastered, adjusted for form difficulty. Some schools still call it your “raw score” in their grading policies, which adds to the confusion.
Your percentile shows how your EPC compares with a national reference group of students who took the same subject exam. Schools usually look up your percentile in NBME’s norm tables.
Your score describes how much you know. Your percentile describes how you compare with others. A score of 80 and the 80th percentile are two very different things.
| Term | What it means |
| Raw score | Your direct test performance on scored questions |
| EPC | NBME’s equated measure of the content you’ve mastered |
| Percentile | Your position compared with a reference group |
| Clerkship grade | The grade your medical school assigns |
You may also see older terms online, such as “scaled score,” “standard score,” or “three-digit score.” NBME moved clinical science subject exams from a scaled score to EPC scores in 2015. Three-digit scores belong to USMLE exams such as Step 2 CK, not to current shelf reports.
Before looking at the numbers, keep one thing in mind. Percentile relationships vary by subject, by year, and by testing period. The tables below are reference information, not a universal NBME conversion formula.
NBME publishes its official norm tables to schools through its faculty portal. What students usually see online are tables that individual schools publish. For this guide, we use a public example from the University of Tennessee Health Science Center (UTHSC) College of Medicine, 2025–2026 academic year. You can view the original UTHSC 2025–2026 M3 Grading Cutoffs.
How to read these tables:
These are one school’s cutoffs for one academic year. Your school’s numbers may be different.
The IM shelf is one of the most important shelves for students heading toward Step 2 CK, because so much of Step 2 CK is internal medicine.
UTHSC 2025–2026 Internal Medicine cutoffs (school-specific):
| Percentile benchmark | Q1 | Q2 | Q3 | Q4 |
| 75th percentile (Honors) | 78 | 79 | 79 | 80 |
| 55th percentile (High Pass) | 73 | 75 | 75 | 76 |
| 5th percentile (Pass) | 57 | 58 | 59 | 59 |
What this means: In this table, a student needed an EPC of about 78 to 80 to reach the 75th percentile. A score in the low 70s sat a little below the middle of the group. Notice that the cutoffs rise slightly as the year goes on. We’ll explain why later in this guide.
What is a good IM shelf score? Using this example, a score of about 78 to 80 or higher was in the top quarter. A score around the mid-70s was near the middle. Always check your own school’s table before judging your result.
UTHSC 2025–2026 Surgery cutoffs (school-specific):
| Percentile benchmark | Q1 | Q2 | Q3 | Q4 |
| 75th percentile (Honors) | 77 | 79 | 79 | 79 |
| 55th percentile (High Pass) | 73 | 74 | 75 | 75 |
| 5th percentile (Pass) | 57 | 59 | 60 | 60 |
What this means: The Surgery numbers look a lot like Internal Medicine. That isn’t a coincidence. The Surgery shelf tests a lot of medicine, such as fluids, electrolytes, and perioperative care. Early in the year, a 77 reached the 75th percentile. Later in the year, students needed a 79.
UTHSC 2025–2026 Pediatrics cutoffs (school-specific):
| Percentile benchmark | Q1 | Q2 | Q3 | Q4 |
| 75th percentile (Honors) | 82 | 83 | 83 | 83 |
| 55th percentile (High Pass) | 77 | 78 | 79 | 79 |
| 5th percentile (Pass) | 62 | 63 | 64 | 64 |
What this means: Pediatrics cutoffs sit a few points higher than IM and Surgery. A score of 80 on Peds landed between the 55th and 75th percentile in this table. The same 80 on the IM shelf was at or above the 75th percentile in every quarter. This is a clear example of why one generic conversion chart doesn’t work.
UTHSC 2025–2026 Psychiatry cutoffs (school-specific):
| Percentile benchmark | Q1 | Q2 | Q3 | Q4 |
| 75th percentile (Honors) | 88 | 88 | 89 | 89 |
| 55th percentile (High Pass) | 85 | 85 | 86 | 86 |
| 5th percentile (Pass) | 73 | 74 | 74 | 74 |
What this means: Psychiatry has the highest cutoffs of all the subjects in this table. Students tend to score high on the Psych shelf, so a high number doesn’t stand out as much. For example, an 80 on Psych fell between the 5th and 55th percentile. A score in the low 70s could fall below the passing percentile.
If you see a Psych score that looks great next to your IM score, check the percentile before you celebrate or worry.
UTHSC 2025–2026 OB/GYN cutoffs (school-specific):
| Percentile benchmark | Q1 | Q2 | Q3 | Q4 |
| 75th percentile (Honors) | 83 | 84 | 84 | 84 |
| 55th percentile (High Pass) | 80 | 80 | 80 | 80 |
| 5th percentile (Pass) | 65 | 65 | 66 | 66 |
What this means: OB/GYN cutoffs are also fairly high. In this example, an EPC of 80 placed a student at about the 55th percentile. To reach the 75th percentile, a student needed an 83 or 84.
UTHSC 2025–2026 Family Medicine cutoffs (school-specific):
| Percentile benchmark | Q1 | Q2 | Q3 | Q4 |
| 75th percentile (Honors) | 79 | 80 | 80 | 80 |
| 55th percentile (High Pass) | 75 | 75 | 75 | 75 |
| 5th percentile (Pass) | 61 | 62 | 62 | 62 |
What this means: Family Medicine sits close to IM, which makes sense. The FM shelf covers a wide range of adult and outpatient medicine. In this table, a 75 was at the 55th percentile, and a 79 or 80 reached the 75th percentile.
UTHSC 2025–2026 Neurology cutoffs (school-specific):
| Percentile benchmark | Q1 | Q2 | Q3 | Q4 |
| 75th percentile (Honors) | 85 | 85 | 85 | 85 |
| 55th percentile (High Pass) | 81 | 82 | 82 | 82 |
| 5th percentile (Pass) | 66 | 67 | 67 | 68 |
What this means: Neurology cutoffs are higher than IM and Surgery but lower than Psychiatry. In this example, an 85 was needed for the 75th percentile in every quarter.
Here is how an EPC of 80 compared across subjects in the UTHSC Q1 table:
| Subject | Where an 80 fell (Q1) |
| Surgery | Above the 75th percentile |
| Internal Medicine | Above the 75th percentile |
| Family Medicine | Above the 75th percentile |
| Pediatrics | Between the 55th and 75th percentile |
| OB/GYN | At about the 55th percentile |
| Neurology | Between the 5th and 55th percentile |
| Psychiatry | Between the 5th and 55th percentile |
Same number. Very different meaning.
There isn’t one meaningful average for all shelf exams. Each subject has its own score distribution, and NBME provides separate norms for each exam.
You can get a rough idea from the tables above. In the UTHSC 2025–2026 example, the score at the 55th percentile ranged from 73 (IM and Surgery, Q1) to 86 (Psychiatry, Q3 and Q4). The middle of the group sits just below those numbers. So an “average” score on the Surgery shelf looks very different from an “average” score on the Psych shelf.
It depends on the subject. In the UTHSC example, a 70 on Internal Medicine or Surgery was comfortably above the passing percentile but below the middle of the group. On Psychiatry, a 70 fell below the 5th-percentile passing cutoff. So a 70 can mean “passing but below average” on one shelf and “not passing” on another.
A 75 was close to the middle on IM, Surgery, and Family Medicine in the UTHSC table. In Q1, it reached the High Pass cutoff for both IM and Surgery. On Pediatrics and OB/GYN, it was below the middle. On Psychiatry, it was only just above the passing cutoff.
On IM, Surgery, and Family Medicine, an 80 was at or above the 75th percentile in the UTHSC table. That’s a strong result. On Pediatrics and OB/GYN, it was a solid middle score. On Neurology and Psychiatry, it was below the middle.
On most subjects, an 85 is a strong score. In the UTHSC table, it cleared the 75th percentile for IM, Surgery, Family Medicine, Pediatrics, and OB/GYN, and it matched the Neurology Honors cutoff. On Psychiatry, an 85 was at about the 55th percentile.
The lesson is the same each time. Always read your score with the norms for your subject and, if your school uses them, your testing period.
There is no single passing percentile that every medical school uses. Each school sets its own grading policy.
Many schools set the passing cutoff at a low percentile. At UTHSC, the passing cutoff for 2025–2026 was the 5th percentile. At Texas Tech University Health Sciences Center, school policy describes the 10th percentile as the minimum passing level for clerkship NBME exams. Other schools use a fixed EPC score instead of a percentile.
This also varies by school. UTHSC used the 75th percentile as the minimum shelf score to be eligible for Honors in 2025–2026. Some schools use a higher bar, and some use a different method entirely. Also remember that a strong shelf score often makes you eligible for Honors. Your clinical evaluations and other components may still decide the final grade.
A few reasons:
Your school’s clerkship handbook or grading policy is the only reliable source for your cutoffs.
No. This is one of the most common mistakes students make.
An EPC of 80 on one subject should not be read with the same percentile as an EPC of 80 on another subject. Each exam has different content, a different student group, and a different score spread.
As the tables above show, an 80 was a top-quarter score on Surgery but a below-middle score on Psychiatry in the same school’s table. If someone tells you “80 is the 70th percentile,” ask which subject, which year, and which testing period they mean.
It can, depending on how your school grades.
As students move through the year, they finish more rotations and gain more clinical experience. NBME’s score interpretation guide explains that it provides quarterly norms because, on some subject exams, students of the same ability tend to score higher when they test later in the academic year.
Some schools use these quarterly norms. The UTHSC example shows this clearly. On the IM shelf, the Honors cutoff was 78 in Q1 and 80 in Q4. On Surgery, it rose from 77 to 79.
Not because the exam itself gets harder. The exam is built and equated the same way. What changes is the comparison group. Later in the year, students taking the exam have more clinical experience, so scores in the reference group tend to be higher. If your school uses quarterly norms, you may need a slightly higher score later in the year to reach the same percentile.
The fair side of this system is that early-rotation students are not penalized for having less experience.
Your shelf score is usually one part of your clerkship grade. Other parts often include:
Common grade categories include Pass, High Pass, and Honors. Some schools also use Fail or a conditional grade if a student needs to retake the shelf.
Your final clerkship grade then appears on your medical school transcript and is summarized in your Medical Student Performance Evaluation (MSPE). NBME itself advises that subject exam scores should not be used alone to decide grades, but together with other measures of student performance.
The most important point: your school’s grading policy decides how much your shelf counts. At one school, the shelf may be a large share of your grade. At another, it may be a minimum threshold you simply need to clear.
In general, individual shelf scores are not sent to residency programs as part of your standard USMLE transcript.
Through ERAS, programs receive several separate documents, including:
So programs usually see the effect of your shelf scores through your clerkship grades and MSPE, not the raw number itself.
That said, schools differ in how they write MSPEs and transcripts. Some may mention shelf performance in more detail. It’s best not to assume that programs will “never” see anything related to your shelf. Ask your school how clinical performance is reported.
Do Shelf Exams Matter for Step 2 CK?
Yes, in a practical sense. Shelf exams and Step 2 CK both test clinical reasoning. They use similar question styles, and the content overlaps a lot. Studying well for each shelf builds the foundation you’ll need for Step 2 CK.
Research also shows a link between shelf performance and Step 2 CK scores. A study in Academic Medicine (2012) found a correlation of 0.77 between students’ combined subject exam scores and their Step 2 CK scores at one medical school. A larger 2022 study in the Journal of General Internal Medicine found that Step 1 performance and the NBME Medicine subject exam together explained about 60% of the variation in Step 2 CK scores.
But association is not a guarantee. The same 2022 study found that simply taking more NBME subject exams was not linked to better Step 2 CK scores. What matters is what you learn, not how many exams you sit.
A strong shelf score is a good sign. A weak one is a signal to adjust your study plan. Neither one decides your Step 2 CK score on its own.
Our USMLE program helps IMGs turn shelf and clerkship performance into a structured Step 2 CK plan that targets the systems costing you points.
This question matters a lot for international medical graduates, and the answer needs some care.
Shelf exams are mainly part of medical school clerkship assessment. NBME sells them to schools, and schools give them in a secure, proctored setting. They are not like USMLE exams, where an individual registers directly and receives an official score.
This means:
For most IMGs, the real value of shelf-style content is Step 2 CK preparation, not the shelf score itself.
Many students take NBME practice forms before their shelf. It’s easy to mix up the different numbers you see during preparation. Let’s sort them out.
This is your EPC from the real subject exam, given by your school. It is the only number your school uses for grading.
NBME’s practice forms for shelf exams are the Clinical Science Mastery Series. These use a different score scale. According to NBME’s sample score report, scores are scaled to a mean of about 20 and a standard deviation of about 3, with total scores ranging from 1 to 30.
Each score report includes a conversion table that translates your practice score into an approximate subject exam score. For example, NBME’s Psychiatry sample report shows that a practice score of 20 approximates a Psychiatry subject exam score of 76.
However, NBME also clearly states that it does not guarantee that your practice score will predict your real shelf performance. It cautions students against using the practice score to make predictions.
Your UWorld or other question bank percentage is your percent correct on that bank’s questions. It is useful for tracking your progress. It is not an NBME score, and it cannot be turned into an official shelf percentile.
Counting how many questions you got right on a practice form gives you a rough raw score. It doesn’t account for form difficulty or scoring rules, so it’s only a loose estimate.
The key rule: don’t convert a practice-form percentage into an official shelf percentile using a generic online calculator. Many of these calculators describe their own results as illustrative, not official. Use the conversion table in your NBME score report, treat it as approximate, and then check your school’s norms.
Not sure which practice forms to use next? See which NBME is best for Step 1.
Your shelf results can guide your Step 2 CK preparation. Here is a simple framework based on where your percentile usually falls.
Focus on building a stronger base:
Focus on steady improvement:
Focus on keeping what you’ve built:
No shelf percentile guarantees a specific Step 2 CK score. Use your shelf results as feedback, not as a prediction.
See the proven Step 2 CK study strategies and resources IMGs use to close content gaps and lift their readiness before test day.
Shelf exams and Step 2 CK are related, but they serve different purposes.
| Feature | Shelf Exam | Step 2 CK |
| Purpose | Clerkship assessment | Medical licensing exam |
| Administered by | NBME Subject Exam program, through your school | USMLE program |
| Score format | EPC score, with percentile from norms | Three-digit USMLE score |
| Subject scope | One clinical subject | Broad clinical medicine |
| Used for clerkship grade | Yes | No |
| Used for residency application | Indirectly, through grades and MSPE | Directly, through the USMLE transcript |
Your EPC is close to a percent-correct score, but it is adjusted for form difficulty. It also leaves out field-test questions and any questions removed from scoring.
Each subject has its own norms. An 80 on Surgery and an 80 on Psychiatry can land in very different percentiles.
Online calculators are estimates. Only NBME norms and your school’s grading table can tell you your actual percentile.
A classmate’s 82 on OB/GYN in Q4 can’t be compared directly with your 78 on IM in Q1. Different subjects and different testing periods use different comparisons.
Programs usually see your clerkship grades and MSPE, not your individual shelf scores.
One weak shelf is feedback, not a final verdict. Scores also have some measurement error, often a few points. Many students improve a lot between clerkships and Step 2 CK.
NBME reports your shelf performance as an Equated Percent Correct (EPC) score. This estimates the percentage of content you’ve mastered, adjusted for differences in exam form difficulty. Your school then compares your EPC with NBME norms to find your percentile and assign your grade.
Not in the traditional sense. NBME equates scores across exam forms, and your percentile compares you with a national group. Your school’s grading cutoffs are a separate step.
There is no single average across all subjects. Each shelf has its own distribution. In one school’s 2025–2026 table, the 55th-percentile score ranged from 73 on IM and Surgery to 86 on Psychiatry.
It depends on the subject. In the UTHSC 2025–2026 example, a 70 on IM or Surgery fell between the 5th and 55th percentile. On Psychiatry, it fell below the 5th percentile.
On IM, Surgery, and Family Medicine, a 75 was near the middle in the UTHSC table. On Pediatrics, OB/GYN, Neurology, and Psychiatry, it was below the middle.
On IM, Surgery, and Family Medicine, an 80 was at or above the 75th percentile in the UTHSC table. On OB/GYN, it was around the 55th percentile. On Psychiatry, it was below the middle.
An 85 was above the 75th percentile on most subjects in the UTHSC table. On Psychiatry, it was around the 55th percentile.
Many students consider the 75th percentile or higher a strong result, because some schools use it as an Honors threshold. A score around the 50th to 55th percentile is roughly average. Your school’s cutoffs define what “good” means for your grade.
It depends on your school. Common passing cutoffs are low, such as the 5th or 10th percentile, but some schools use a fixed score instead.
This varies by school. One common example is the 75th percentile, but some schools set a higher bar. A strong shelf score may only make you eligible for Honors, not guarantee it.
The IM shelf uses the same EPC system as other NBME clinical subject exams. Your EPC is compared with Internal Medicine norms to find your percentile.
The 75th percentile or higher is a strong IM result. In the UTHSC 2025–2026 table, that meant an EPC of about 78 to 80, depending on the quarter.
The 75th percentile or higher is strong. In the UTHSC 2025–2026 table, that meant an EPC of about 79 to 80.
The 75th percentile or higher is strong. In the UTHSC 2025–2026 table, that meant an EPC of about 83 to 84.
No. Each subject has its own norms. The same EPC can fall in very different percentiles depending on the subject.
Not reliably on your own. You don’t know which questions were scored, and your EPC is adjusted for form difficulty. Use your official EPC and your school’s norm table.
Only approximately. The Clinical Science Mastery Series score report includes a table that converts your practice score to an approximate subject exam score. NBME cautions that this is not a prediction of your real result.
Usually not directly. Programs typically see your clerkship grades on your transcript and a summary of your performance in your MSPE.
Yes. They test similar clinical reasoning, and research shows a link between shelf scores and Step 2 CK scores. Still, a shelf score does not guarantee any specific Step 2 CK result.
Usually only through a medical school or clinical program that gives the exam. IMGs cannot typically register for an official shelf on their own. They can, however, buy NBME Clinical Science Mastery Series practice forms.
A shelf score only makes sense when you know what it is and what it’s being compared with. Your EPC tells you how much content you’ve mastered, adjusted for exam difficulty. Your percentile tells you where you stand compared with other students. Your school decides what that means for your grade.
The same number can mean very different things across subjects and testing periods. So before you celebrate or worry, check three things: the subject, your school’s norm table, and your testing period. Most importantly, use your shelf results as feedback. Each shelf is a chance to find your weak spots and build the clinical reasoning you’ll need for Step 2 CK and beyond.
Not sure which practice forms to use next? See which NBME to take for Step 2 CK.
Shelf scores are only one piece of the puzzle. As an IMG, you also need a clear plan for Step 2 CK, clinical experience, and your residency application. IMG Helping Hands guides international medical graduates through every stage of the USMLE pathway, so you always know your next step.
Shelf scores are one signal. As an IMG, you need the whole picture pointing at residency. We help you connect the dots:
Articles published by IMG Helping Hands are prepared by our team using information from direct experience, publicly available resources, and educational references. AI tools may be used to assist with drafting, proofreading, and formatting; however, all content undergoes review and approval before publication.
The information provided is intended for educational purposes only. Requirements, policies, and processes may change over time. Readers should consult official sources for the most current information.
Get our most valuable tips right inside your inbox, once per month!
