Where to apply

Don't swim upstream

Most large programs screen applications before a human reads them — by USMLE Step 2 CK score, year of graduation, U.S. clinical experience, visa type sponsored, and ECFMG certification. If your application trips a program's hard filter, no one there ever reads your story, however good it is. In NRMP's 2024 Program Director Survey, the factors the most program directors cite for deciding whom to interview are a Step 1 pass (90%), the MSPE (85%), and specialty letters (84%) — the screen is real and it is early.

Know your own gated metrics first — your Step 2 CK band, how many years since graduation, your USCE, which visa you need. The red-flags guide maps which of these are correctable and which you reframe; this page is about aiming the list at programs those metrics can actually reach.

The targeting tools

Two free tools turn "realistic" from a guess into data:

• Residency Explorer (AAMC and partners, free with an AAMC account). For each program it shows the characteristics of applicants who were invited to interview and matched — including interview-invitation rates broken out by applicant type (MD, DO, U.S. IMG, and non-U.S. IMG) and the Step 2 CK score range of invited applicants. Compare yourself, program by program. It is explicit that it does not predict where you'll be invited or tell you where to apply — but it is the best available read on whether you're in a program's range.
• FREIDA (the AMA's database of 13,000+ ACGME-accredited programs — free to search and filter; some expanded program detail requires AMA membership). Filter by the things that decide IMG eligibility — whether the program sponsors J-1 and/or H-1B visas, and the percentage of IMGs among its current residents. A program that doesn't sponsor your visa is not a program you can rank, so filter those out before anything else.

Compatible and realistic — with room to reach

Aim for deeply compatible, desirable programs that your metrics can realistically reach. "Realistic" doesn't mean timid — a reach is fine, and healthy, as long as the reach is plausible on your gated numbers rather than a fantasy a filter will erase. The list you want is mostly programs squarely in your range, salted with a few genuine reaches and anchored by programs you're confident about (your "safeties").

Weigh compatibility on what actually predicts a good three years and a good match: a program's track record with IMGs and with your visa, its case mix and location, and honest fit. Score candidates as you research them — the Program Evaluation Checklists exist so this is a structured comparison, not a vibe.

How many programs

More is not linearly better. AAMC's Apply Smart research established a point of diminishing returns — a number beyond which each additional application barely moves your probability of landing a position — and that point is specific to your specialty, your exam performance, and your applicant type (the data breaks out U.S. IMGs and non-U.S. IMGs separately). IMGs generally do need to apply broadly, but "broadly" has a ceiling past which you're spending money for almost nothing. And it isn't cheap: for 2027, ERAS charges $11 per program for the first 30 in a specialty and $30 each beyond that — so a 60-program list runs about $1,230 before ECFMG and other fees (confirm current pricing on the AAMC schedule).

Where to signal

Program signals let you flag genuine interest to a limited number of programs — for the 2027 season Internal Medicine gets 15 (3 gold and 12 silver), and counts vary widely by specialty and roll each cycle. Spend them the way the whole list should be built: on realistic-fit programs you genuinely want, not on reaches where you don't clear the gates. A signal is most valuable where you're a plausible candidate and interest could tip the decision — and close to wasted on a program that will filter you out anyway. How to value a signal, why not signaling now says something, how to split golds from silvers, and the full per-specialty table are in the dedicated guide: Preference Signaling.

Applying to a second specialty

Many applicants apply in two specialties in the same season — commonly Internal Medicine paired with a related field such as Neurology or Family Medicine. Mechanically, it is one ERAS application: you apply to programs across both specialties, assign a specialty-appropriate personal statement to each program, and assign letters per program. The honest trade-offs:

• It widens the pool you're eligible for. Applying in a second field opens programs that would otherwise be closed to you — which is why applicants do it. But be honest about the evidence: no NRMP or AAMC figure shows dual-applying raises match odds, and the nearest available measure — how many specialties matched versus unmatched applicants ranked — is confounded and even leans the other way, because unmatched applicants often add a specialty out of worry. Treat it as an eligible-pool widening whose net effect is unmeasured, not a numerical edge.
• It costs more. For 2027, ERAS charges $11 per program for the first 30 in a specialty and $30 each beyond that, and the pricing resets with each new specialty — so a second specialty starts again at the $11 tier, on top of the first.
• It adds complexity, and a question to answer. You'll write and manage a distinct personal statement per specialty, and you should be able to explain the choice coherently — programs can read a scattered dual application as a lack of commitment. Apply to a second specialty because you'd genuinely train in it, not as a hedge you can't defend.

Geography & visa

Visa-sponsoring, IMG-friendly programs are not spread evenly across the country. NRMP's results by state, specialty, and applicant type show where IMGs actually fill — a more honest guide than reputation. Weigh location for the same real-life reasons a second look exists (cost of living on a resident salary, community, getting around), but for an IMG the first filter is always practical: does the program sponsor the visa you need?

One longer-range note: the after-residency visa picture (J-1 waivers through routes like Conrad 30) is geographic too, and can make a region that's convenient now inconvenient in five years — but that belongs to career planning, not this list. See J-1 waiver paths when you get there, and ECFMG & the Pathways for the certification and J-1 sponsorship mechanics.