Preference signaling

Why signals exist

Programs receive far more applications than anyone can read closely, and the application itself stopped being a credible sign of interest the moment it became as easy to send a hundred as to send ten. A program signal restores that credibility because it is limited: a program that receives one knows you chose it over the dozens you didn't. That is the AAMC's stated purpose — a way to "express interest in individual programs at the time of application," so a program can find the applicants who are both qualified and genuinely interested.

Two things follow, and the rest of this page rests on them. Programs want signals, because a signaled applicant makes their interview slate more efficient and lowers the risk of an unfilled class. And because a signal is scarce, its absence carries information too.

What a signal is worth

A signal's value is the product of two things: how much you want the interview it might produce, and how much the signal changes your odds of getting that interview. Optimize those two, and nothing else.

Why the interview, and not the match? Because the invitation is the scarce step. Programs rank the large majority of the applicants they interview — in NRMP's 2022 Program Director Survey the average program reported interviewing 111 applicants and ranking 100 of them — so an invitation means the program has already decided you are worth ranking, and a signal's leverage is almost entirely upstream, at the invitation. Some programs do look at signals again when they rank — about one program director in six told NRMP's 2024 signaling survey that signals affected who made the rank list, even though the AAMC instructs programs to use them "during the interview selection process only" — but don't allocate around that. It is a downstream effect you can't control, and optimizing for it makes you spend badly at the step that counts.

The second term is the one applicants get wrong. For each program, ask two questions. Would they interview me anyway? If yes, a signal there buys nothing you didn't already have. Would they interview me even with a signal? If no — if your metrics trip their filter — it buys nothing either. The signal earns its keep in the middle: programs where you are clearly qualified but your interest is the open question — you look like a safety, every address in your file is on the other coast, or they simply have more qualified applicants than interview slots and need a reason to pick you.

Five principles

1. Signals don't change your qualifications. A program that would screen your application on Step 2 CK, year of graduation, visa, or U.S. clinical experience will still screen it; the signal is never read. Signaling only moves you from "qualified, but is this applicant interested?" to "qualified and interested." Spend signals where you already clear the gates — that is the whole argument of Where to Apply, and Red Flags & Filtered Metrics tells you which gates you can still fix.
2. You only win if you get something you want. Before you touch a signal, list the programs you would actually attend, in your genuine order — a private list, not your dean's and not a rankings site's. A strategy that wins interviews at programs you wouldn't go to didn't beat the system; it beat you.
3. Where signals are the norm, no signal is the signal. In a 15-signal specialty, a competitive program receives far more signaled applications than it can interview. It has little reason to open the unsignaled ones — and by not signaling you have told it, in effect, that it is at best your sixteenth choice. At the popular, sought-after programs in any specialty a signal is now the price of a look. This is strategic reasoning, not an AAMC rule, and it is weakest at programs that receive few signals — see the next principle.
4. Zig when others zag. Applicants cluster: in the AAMC's 2024-cycle data, roughly a fifth to a quarter of a specialty's signals went to just 10% of its programs. The AAMC doesn't say which programs those are, but it is not hard to guess — the same big-name and coastal programs — while excellent programs off the beaten path receive few. All else equal, a signal moves your odds more at a program that receives fewer of them. When you are torn between two programs you would be happy at, that is a sound tiebreaker — and the clinical training is often the same.
5. Cover your downside. The cost of not matching dwarfs the cost of matching somewhere that wasn't your first choice. Put at least one signal — and in a competitive or big-signaling specialty, several — on programs where you are clearly well-qualified and would be genuinely content. Then check the whole allocation against your own risk tolerance. Aiming every signal at reach programs is a legitimate choice, but make it with open eyes: programs that would have loved to have you may not look if they were never told you were interested.

Gold and silver

Six specialties use two tiers for the 2027 season — Anesthesiology, Child Neurology, Dermatology, Diagnostic Radiology, Internal Medicine, and Vascular Surgery–Integrated; the exact split is in the table below. The AAMC's own definition is the right starting point: gold signals mark your "most preferred" programs, silver your "preferred" ones. Turn that into an allocation rule, then layer the principles on top:

• Golds are your true top few among programs you realistically clear — not your biggest reaches. A gold on a program that filters you out is a gold wasted, and in a three-gold specialty that is a third of your strongest currency.
• Silvers are the working list: realistic-fit programs you genuinely want, including the ones that cover your downside.
• Use the full allotment. The AAMC advises using every signal you are given. An unused signal is a program that never learned you were interested.

The 2027 table, by specialty

These are the program-signal allotments for the ERAS 2027 season (the 2027 Match) as published by the AAMC. They change most years — confirm them on the AAMC's program-signaling page before you commit. Participation is voluntary at the program level, so a given program in a listed specialty may still have opted out.

The written explanation

Four specialties — Anesthesiology, Physical Medicine & Rehabilitation, Plastic Surgery–Integrated, and Radiation Oncology — require a written explanation of 300 characters or fewer with every program signal, and the signal isn't saved without it. In those fields the explanation can make a signal worth more or less than its face value. Put yourself in the program director's chair and ask what would read as credible. The strong ones share three features:

• Motivating. Programs want residents who will thrive, and thriving usually means support: family or roots in the area, a mentor or rotation at that program, a clinical focus or mission that matches yours, a genuine intent to practice in that region.
• Specific — about you or about the program, never generic praise. "Outstanding clinical training, excellent faculty, and a diverse patient population" describes every program in the country and moves no one. "My research and career interest in resource-limited settings leads toward an academic global-health career, and your program's global-health track builds directly toward it" tells the reader exactly why this program.
• Credible. Say true things. A track that was discontinued two years ago, or a fellowship the program hasn't sent anyone to in years, is exactly the detail a program director will notice — so if you write about the program, do the homework first.

Geographic preferences

Separately from program signals, MyERAS lets you name up to three of the nine U.S. Census divisions as preferred places to train, along with a setting preference (urban through rural), or declare no preference at all. A program sees your geographic preference only if it sits in a division you chose. Three specialties — Orthopedic Surgery, Otolaryngology, and Plastic Surgery–Integrated — have opted out of receiving both geographic and setting preference answers for 2027.

Be honest about what it can do. Its leverage is limited and hard to separate from the connections that are already visible in your file, and in a specialty where applicants have 25 or 30 program signals, a checked box for a region carries little weight next to the signal you didn't send. Where it plausibly helps is in low-signal specialties and at less-subscribed programs. The practical rule: choose the divisions that contain your signaled programs; if three can't cover them, "no preference" is a legitimate answer. And if you write a reason, keep it broadly true across the whole division — a reason that is really about one city can quietly close doors in the rest of it.

For IMGs and DOs

You will hear that signaling "doesn't work" for international graduates, or for DOs. The claim comes from the top-line data, and the data are real: at the median program in every specialty the AAMC reports, an IMG who signals is invited less often than an MD who signals — in some specialties, less often than an MD who didn't signal at all. But that measures which applicants programs prefer, not whether a signal changes your odds. The AAMC's own finding is that "sending a program signal resulted in a higher probability of interview invitation for all participating specialties. However, programs varied widely in how they used signals, and sending a signal did not guarantee an interview invitation." Its 2026-season analysis (data as of May 31, 2026, posted on the AAMC's ERAS statistics page) breaks that out by applicant type, and the IMG lift is real — from a far lower base:

• Internal Medicine: the median program interviewed 24% of the IMGs who sent it a gold signal, 9% of those who sent a silver, and 1% of those who sent none (MDs: 62% / 50% / 11%).
• Family Medicine: IMGs 12% with a signal versus 4% without (MDs: 78% / 49%).
• Anesthesiology: IMGs 15% gold / 3% silver / 0% none (MDs: 67% / 43% / 1%).
• And the honest floor: in Dermatology, Interventional Radiology–Integrated, and Internal Medicine/Psychiatry, the median program interviewed 0% of IMGs with or without a signal.

Those are medians across programs, and the AAMC attaches its own caution: "small DO and IMG sample sizes at the program level," and the analysis "does not control for all factors considered in the selection process." Read them as the shape of the landscape, not your personal odds.

So the right comparison is never you against an MD applicant. In practice IMGs compete mainly against other IMGs, and DOs often against other DOs, for the seats that are realistically open to them — and in that contest a signal is frequently the thing that gets your file opened. So the principles above apply to you unchanged, with three IMG-specific moves:

• Aim the signals at IMG-friendly, visa-sponsoring programs. Filter FREIDA by visa sponsorship and percentage of IMGs; in Residency Explorer, read each program's interview-invitation rates by applicant type — it breaks out U.S. IMGs and non-U.S. IMGs separately — and compare yourself to the applicants it actually invited. A signal to a program that has never sponsored your visa is the purest form of waste.
• Use every signal, on realistic programs. The old strategy of applying to two hundred programs and hoping is precisely the behavior signaling was introduced to blunt. A carefully chosen list of programs where you clear the gates, with the full signal allotment spent on the ones you most want, is a stronger position than an enormous unsignaled spray — and far cheaper (see how many programs).
• Residency Explorer's Interview Likelihood tool is a three-shot resource. You may enter your application profile and receive program-level interview likelihoods, but only three submissions for the whole season — so save them until your application is essentially final rather than war-gaming hypotheticals.

Logistics, for the record: your ECFMG residency token, issued through the MyIntealth applicant portal from June 24, 2026, is what unlocks MyERAS. One token covers every specialty and every program you apply to in the season — see ECFMG & the Pathways.

When signals are submitted

Signals are chosen inside the MyERAS application and travel with it — they are not a separate step after you apply. For programs you have already applied to, you can still add or change signals until three checkpoints in the season: July 15, 2026, September 23, 2026, and December 2, 2026, each at 9:00 a.m. ET. After each checkpoint, the signals to programs you had applied to by then lock in place. For an applicant submitting on the usual schedule, that means the September date is effectively the last moment to re-think an allocation.