Every application cycle, graduate medical education (GME) program directors face a familiar question when screening applications submitted through the Electronic Residency Application Service (ERAS): Where should the numerical filters be set?
Two recent changes have simplified and narrowed that question. USMLE Step 1 is now reported as pass/fail, removing its score variations as a screening variable. USMLE Step 3 scores are rarely available at the time of application, since most applicants take that exam later in training, with the exception of some international medical graduates (IMGs). That leaves USMLE Step 2 Clinical Knowledge (CK) as the only numeric score most programs can realistically use to stratify applicants during initial screening.
The missing half of the filter
Most programs already apply a lower limit on USMLE Step 2 CK, a minimum score below which an application won’t be reviewed further. Far fewer apply an upper limit. That asymmetry is worth reconsidering.
Program directors typically know, or can easily determine, the range of USMLE Step 2 CK scores among their currently enrolled residents. That range reflects the scores of applicants who not only applied and interviewed, but actually matched at the program. It is, in effect, a historical signal of fit, not a ceiling on ability, but a marker of where a given program has actually been competitive.
Real physician voices, twice a week
Free, and one click to unsubscribe.
Applicants whose scores fall far above that historical range are, in practice, unlikely to rank the program highly or to match there. They are typically far more competitive nationally and are pursuing top-tier programs. Inviting them to interview consumes a scarce, non-renewable resource, an interview slot, that could instead go to an applicant whose score profile matches the program’s actual track record of matching residents.
Why this matters for the system, not just one program
If more programs used both a lower and upper bound reflecting their own resident data, the effect would likely compound nationally:
- Fewer mismatched interview offers: Applicants far above a program’s typical range are less likely to occupy a slot they were never going to match.
- More applicants interviewed overall: Freed-up slots could go to applicants who are realistic matches, modestly increasing the total number of applicants who receive at least one interview during the cycle.
- Better-directed effort for high scorers: Applicants denied interviews at programs on this basis could redirect their effort toward programs whose historical residents’ score range aligns with their own scores, arguably a more efficient use of a finite interview season for everyone.
There is also a built-in safety net for high-scoring applicants who take this approach and still end up unmatched. Because they remain highly competitive on paper, they are typically the first applicants absorbed by unfilled programs during the Supplemental Offer and Acceptance Program (SOAP).
Caveats worth naming
This approach is not without risk, and program directors should weigh it carefully:
- Scores a proxy, not a full picture: USMLE Step 2 CK scores correlate only loosely with clinical performance, and an upper limit risks excluding excellent, mission-aligned applicants who simply scored well. Applicant motivations vary. Some high scorers apply broadly or to specific programs for reasons unrelated to score fit, such as geography, family, research interest, or specialty competitiveness, and a rigid upper bound could screen out applicants who would genuinely rank the program first.
- Historical range self-limiting: Anchoring purely to past resident scores may entrench a program’s current profile rather than allow it to grow more competitive over time.
- Equity and compliance review: Any new screening variable should be reviewed against institutional and national guidance, for example AAMC recommendations, to ensure it doesn’t inadvertently disadvantage protected groups or conflict with holistic review principles.
The bottom line
An upper limit on USMLE Step 2 CK scores, informed by a program’s own historical resident data, is not about penalizing high achievement. It’s about aligning interview invitations with realistic matching outcomes. Used thoughtfully, and alongside, not instead of, holistic review, it could make interview season modestly more efficient for applicants and programs alike.
Deepak Gupta is an anesthesiologist. Sarwan Kumar is an internal medicine physician.
