Is USMLE Step 3 Pass/Fail? Scoring and Fellowship Impact

USMLE Step 3 is not pass/fail. It still reports a three-digit numerical score, just as it has for decades. The exam that changed to pass/fail scoring was Step 1, which made the switch on January 26, 2022. Because the two exams get discussed together constantly, and because the Step 1 change dominated medical education news for years, the confusion is understandable. But the distinction matters quite a bit for anyone planning a medical career in the United States.

What Actually Changed With Step 1

Step 1 of the USMLE historically produced a three-digit score that residency programs used heavily to screen applicants. In February 2020, the Federation of State Medical Boards (FSMB) and the National Board of Medical Examiners (NBME) announced Step 1 would shift to pass/fail reporting, and the change officially took effect for exams administered on or after January 26, 2022.¹ The stated goal was to reduce the outsized emphasis on a single licensing exam and encourage residency programs to evaluate applicants more holistically.²

Before the change, Step 1 scores functioned almost like a GPA for medical students: competitive specialties set unofficial score cutoffs, and a low number could shut doors regardless of the rest of an applicant’s profile. The switch to pass/fail meant that students who took Step 1 after January 2022 would receive only a “pass” or “fail” on their transcript, with no number attached. Step 2 Clinical Knowledge (Step 2 CK) and Step 3, however, were left untouched. Both still produce numerical scores.

Why Step 3 Still Carries a Numerical Score

Step 3 sits at a different point in the licensing process than Step 1 or Step 2 CK. Medical students take Step 1 and Step 2 CK during medical school. Step 3 is typically taken during or just after the first year of residency, and it is the final exam a physician must pass before entering fully unsupervised practice. It is designed as a last check on whether a doctor is ready to practice independently, covering clinical decision-making, patient management, and the application of biomedical science in real scenarios.

That role as a safeguard before unsupervised practice is a key reason Step 3 retains its numerical score. Research has shown that physicians who score higher on Step 3 are less likely to face disciplinary action from state medical boards later in their careers.³ That correlation gives regulators a practical reason to keep the score: it provides validity evidence for using Step 3 results in licensure decisions. While correlation is not the same as causation, the link between low scores and later problems in practice gives state medical boards something concrete to weigh when granting licenses.

There is also a regulatory dimension. Each state medical board sets its own passing threshold for licensure, and some states set their bar above the national minimum passing score. A pass/fail system would force all states to use the same binary cutoff. The current numerical score lets individual licensing boards maintain their own standards.

How Step 3 Scores Affect Fellowship Matching

For most residents, the primary function of Step 3 is licensure. You pass, you move on. But for residents pursuing subspecialty fellowships, Step 3 scores play a more meaningful role than many people realize. In ophthalmology, for example, residents who successfully matched into fellowship programs had Step 3 scores averaging about 7 points higher than those who did not match, a difference that was statistically significant even after accounting for Steps 1 and 2.⁴

In infectious diseases fellowship, higher USMLE scores across all three steps were associated with stronger performance on in-training exams, and Steps 2 CK and 3 specifically were linked to greater scholarly productivity and receiving awards during fellowship.⁵ Those findings suggest that Step 3, even though it comes relatively late in training, captures something meaningful about a physician’s clinical reasoning and academic trajectory.

Whether fellowship programs will lean more heavily on Step 3 now that Step 1 is pass/fail remains an open question. The data so far comes from an era when programs could use all three numerical scores together. With Step 1 no longer providing a number, the remaining scored exams, Step 2 CK and Step 3, carry proportionally more weight by default.

The Weight Shift Toward Step 2 CK

The most immediate and visible consequence of Step 1 going pass/fail was not about Step 3. It was about Step 2 CK. Surveys of medical school deans conducted around the time of the transition found that the most common pick for the new number-one factor in residency applications was Step 2 CK. That makes intuitive sense: if residency programs can no longer screen by Step 1 score, they will look to the next available numerical metric, and Step 2 CK is the scored exam most applicants complete before applying to residency.

This shift has created a kind of pressure transfer rather than a pressure reduction. The original rationale for making Step 1 pass/fail was to decrease the stress of a single high-stakes exam, but early reports suggested that students were simply redirecting their anxiety toward Step 2 CK.⁶ Instead of obsessing over a Step 1 score, students now obsess over a Step 2 CK score. The stakes feel just as high because the underlying incentive structure has not changed: competitive programs still need a way to filter thousands of applicants, and a numerical score is the easiest filter to use.

Step 3, because it typically comes after residency matching, has not been the primary beneficiary of this shift. But residents pursuing fellowship are in a different position. For them, Step 3 is completed before the fellowship application process, making it newly relevant as one of the few remaining numerical data points available to fellowship directors.

Which Specialties Feel the Change Most

The Step 1 pass/fail transition has not affected all medical students equally. Students applying to the most competitive residency specialties, those with the highest historical Step 1 score cutoffs, have been left with the most uncertainty. Surveys of medical school deans identified dermatology, neurosurgery, orthopedic surgery, ENT (otolaryngology), and plastic surgery as the specialties where applicants would feel the impact most acutely.

In those fields, Step 1 scores once served as an efficient (if blunt) screening tool. A program receiving hundreds of applications could quickly narrow the pool by setting a minimum score. Without that number, programs have had to lean on other metrics: Step 2 CK scores, research output, letters of recommendation, clerkship grades, and institutional prestige. This is what the NBME and FSMB envisioned as “holistic review,” but in practice, it has raised concerns that less transparent factors, like which medical school an applicant attended or which mentors they had access to, may play a larger role.

There is also a diversity question. Roughly six in ten medical school deans surveyed did not believe the pass/fail change would adequately address future diversity in medicine. The worry is that without a standardized numerical score available to everyone, applicants from less well-known medical schools or those without strong professional networks may lose one of the few objective metrics that could help them stand out.

The Well-Being Question

A major selling point for the Step 1 scoring change was student well-being. Step 1 preparation had become an enormous stressor, with many students taking weeks or months of dedicated study time, spending thousands of dollars on commercial prep courses, and reporting high levels of anxiety tied to the exam. The hope was that pass/fail scoring would ease some of that burden.

Early evidence, though, paints a more complicated picture. While the announcement was partly intended to decrease stress associated with a single high-stakes licensing exam, it introduced new worries: students reported anxiety about how they would differentiate themselves without a Step 1 score, uncertainty about what residency programs would value instead, and concern that the change could disadvantage them relative to previous cohorts who had numerical scores to show.⁷ The unanticipated secondary consequences of the change have been a running theme in medical education discussions since the announcement.

For Step 3 specifically, the well-being implications are less dramatic. Step 3 was never the major stress point in the licensing sequence because it comes after the residency match, the career milestone most students are anxious about. Failing Step 3 is rare and has serious consequences for licensure, but the intense score-chasing culture that surrounded Step 1 never applied to Step 3 in the same way. Most residents aim to pass comfortably rather than maximize their number.

What Step 3 Actually Covers

Step 3 is a two-day exam, the longest in the USMLE sequence. Day one consists of multiple-choice questions organized into blocks, covering foundational clinical science and the application of biomedical knowledge to patient management. Day two adds computer-based case simulations (CCS), where examinees manage virtual patients over simulated time periods, making decisions about diagnosis, ordering tests, prescribing treatments, and monitoring outcomes. The CCS component is unique to Step 3 and is designed to test real-time clinical decision-making rather than just factual recall.

The exam emphasizes ambulatory (outpatient) medicine more than its predecessors. While Step 1 focuses on basic science and Step 2 CK on clinical knowledge across settings, Step 3 is geared toward the kind of independent patient management a physician would do in an office practice or clinic. It also tests understanding of population health, preventive medicine, and the interpretation of clinical studies, reflecting the expectation that a licensed physician can evaluate evidence and apply it to patient care.

The minimum passing score for Step 3 is 198, and scores can range up to 300. Most examinees pass on the first attempt, though international medical graduates historically have lower first-attempt pass rates than graduates of U.S. allopathic and osteopathic programs. For residents taking the exam, the typical preparation period is much shorter than what students invest in Step 1, often a few weeks of review during residency rather than a dedicated study block.

Could Step 3 Ever Go Pass/Fail

There is no active proposal to change Step 3 to pass/fail reporting. The rationale for the Step 1 change was heavily tied to how that specific score had been co-opted by the residency selection process in ways that were never intended. Step 1 was designed as a licensing exam, not a competitive ranking tool, but it had become the latter. Step 3, by contrast, has not been weaponized in the same way. It functions closer to its original purpose: a check on readiness for unsupervised practice.

The evidence linking Step 3 performance to real-world outcomes, such as the association between lower scores and higher rates of disciplinary action, also provides a practical argument for keeping the numerical score.⁸ If a score correlates with measurable differences in physician behavior, regulators have reason to maintain it as a data point. A binary pass/fail would collapse that gradient and potentially remove a signal that helps state licensing boards identify physicians who may need additional oversight.

There is also the fellowship angle. If Step 3 went pass/fail, fellowship programs would lose one of the few remaining standardized, scored metrics for comparing applicants across different residency programs. Given that the Step 1 change already eliminated one such metric, removing another could accelerate the trend toward less transparent and potentially less equitable selection processes.

Step 3 for International Medical Graduates

For international medical graduates (IMGs), the USMLE sequence carries particular weight because it is often the primary credential that U.S. residency and licensing systems evaluate. Step 3 has a unique significance for IMGs because some states allow candidates to take it before entering a residency program, while others require completion of a certain amount of graduate medical education first. This state-by-state variation matters because passing Step 3 before residency can strengthen a visa application (particularly for J-1 visa waivers and H-1B sponsorship) and demonstrate commitment to practicing in the United States.

Because Step 1 no longer provides a numerical score, IMG applicants who took the exam after January 2022 cannot use a high Step 1 score to distinguish themselves from graduates of U.S. medical schools. This makes Step 2 CK and, for those applying to fellowships, Step 3 proportionally more important as differentiators. An IMG with a strong Step 3 score has a concrete data point to present to fellowship directors, whereas before the Step 1 change, that number might have been overshadowed by the Step 1 score.

The stakes of failing Step 3 are also higher for IMGs than for most U.S. graduates. A failed attempt can complicate visa timelines, delay licensure, and raise red flags with residency programs. While the overall pass rate for Step 3 is high among all examinees, the gap between U.S. graduates and IMGs on first-attempt pass rates has been a longstanding feature of USMLE data, and it underscores the importance of adequate preparation even though the exam is less culturally prominent than Step 1 or Step 2 CK.

Practical Timing and Strategic Considerations

Most U.S. medical graduates take Step 3 during their first year of residency (PGY-1), though the exact timing varies by specialty and state requirements. Some residency programs encourage taking it early in intern year before clinical duties become too demanding, while others are more flexible. In states that require Step 3 for a full medical license (as opposed to a training permit), the timing can affect whether a resident can moonlight, which often requires a full unrestricted license.

For residents who plan to pursue fellowship, taking Step 3 sooner rather than later has a strategic benefit: it ensures the score is available on applications. Fellowship programs in fields like ophthalmology, where matched applicants had measurably higher Step 3 scores, may weigh the number more heavily than programs in less competitive subspecialties.⁹ Residents who delay Step 3 until after fellowship applications may miss the window entirely.

The financial cost of Step 3 is also worth noting. As of recent fee schedules, the exam costs over $900, making it the most expensive individual step in the USMLE sequence. Combined with the cost of any preparation materials, the exam represents a meaningful expense for residents who are already managing medical school debt and earning a trainee salary. Unlike Step 1, where the pass/fail change reduced the incentive to invest heavily in prep, Step 3’s continued numerical scoring means that residents aiming for competitive fellowships still have a reason to prepare seriously and pursue the highest score they can manage.