The Couples Match in residency is often described as a tool for keeping two applicants in the same place. For international medical graduates, it is something more structural: a constraint multiplication problem in which every filter that already operates on one applicant must now clear for both, applied to every paired combination on the rank list.
Coupled rank lists do not change a program’s individual selection criteria. They multiply the structural filters each partner must clear. For IMGs, where SUVY filters and visa sponsorship already narrow the program pool, paired matching makes the math harder, not easier.
The Couples Match has produced match rates above 90 percent every year since 1984. That headline number is real, but it is not the full picture. Beneath it sits a more important fact for IMGs: roughly 63 percent of participating couples are U.S. MD seniors, which means non-USMD couples — including IMG-IMG and IMG-USMG pairings — operate in the minority of the dataset, against a different set of structural constraints than the majority population the headline rate reflects.
The Couples Match is an option offered by the National Resident Matching Program (NRMP) that allows two applicants to link their rank order lists into pairs of program combinations. Rather than each applicant submitting an independent list of preferred programs, a coupled pair submits a list of paired programs — combinations in which Partner A is matched at Program 1 and Partner B is matched at Program 2.
The matching algorithm then attempts to place both applicants at the highest-ranked pair on their list where each partner has been offered a position. If no offer is available for either partner at the top-ranked pair, the algorithm moves down the list until it finds the highest-ranked combination where the placement works for both partners simultaneously.
According to the NRMP, “any two applicants can choose to participate in an NRMP Match as a couple.” The structure is not limited to spouses or romantic partners. Friends, classmates, colleagues, or family members can register together provided both are individually eligible for placement in a U.S.-based residency program.
A critical clarification. The Couples Match does not change how programs evaluate applicants. Each partner is screened, reviewed, and offered positions independently, based on their own qualifications. Coupling affects only the algorithm’s placement logic at the end of the process — not the program’s assessment of either applicant during application or interview review.
In IMGPrep’s Match Funnel framework, the Match is a multi-stage filtering process. Each stage applies a different filter, and each filter uses different inputs. The Couples Match operates formally at the final stage — rank order list submission and algorithmic match — but its constraints propagate backwards through every earlier stage of the funnel.
This is the structural insight most Couples Match guides miss. The decision to couple is not a Stage 6 decision — it is a Stage 1 decision that shapes the entire application strategy. By the time two applicants are submitting paired ROLs in February, the program pool has already been determined by where each partner applied, where each partner interviewed, and where the joint geography of those two interview lists overlaps.
IMGPrep’s SUVY framework — Scores, US clinical experience, Visa, Year of graduation — describes the four structural variables programs use to filter IMG applications before any holistic review begins. For an individual IMG applicant, the strategic task is identifying programs where their SUVY profile clears the program’s thresholds.
For a couple, the math changes. The viable program pool for the couple is the intersection — not the union — of the two individual viable pools. A program is only viable for the couple if both partners’ SUVY profiles clear that program’s thresholds independently. The couple cannot rank a program where Partner A is competitive but Partner B is not, because the algorithm requires both partners to receive an offer at that pair.
This intersection problem becomes severe when the two partners have asymmetric profiles. A couple in which one partner has Step 2 CK 260 and the other has Step 2 CK 220 cannot rank academic programs where the higher-scoring partner is competitive but the lower-scoring partner is below threshold. They are forced toward programs where both clear — which typically means lower-tier programs relative to the higher-scoring partner’s individual viable pool.
For IMG-IMG couples, this compounding is most acute. Both partners face the structural Stage 1 filters of visa status, year of graduation, ECFMG certification, and program-specific IMG-friendliness. The viable program pool for an IMG-IMG couple is often a small fraction of either partner’s individual viable pool. Strategic program selection — not algorithmic luck — is what makes the Couples Match work for IMGs.
For IMG couples, visa status is one of the most underappreciated filters in the paired program pool. Programs that sponsor J-1 visas may not sponsor H-1B; programs that sponsor H-1B may not accept J-1; some programs sponsor neither. When two IMG partners require different visa types, the program pool narrows to programs that sponsor both.
Common configurations IMG couples should plan around:
The 2026 NRMP data made the visa-as-first-order-filter point empirical. In the 2026 Main Residency Match, non-U.S. IMGs requiring visa sponsorship matched at 54.4 percent — a five-year low — while permanent-resident non-U.S. IMGs matched at 67.9 percent — a five-year high. The 13.5-point gap is the clearest evidence yet that visa status functions as a binary structural filter, not a soft preference. For IMG couples, this gap compounds: a couple where both partners require sponsorship faces the smallest program pool of any IMG configuration.
Once both partners have completed interviews, the central strategic task of the Couples Match is constructing the paired ROL. The mechanics are precise:
That last point is the structural risk of the Couples Match. A couple that ranks too narrowly — for example, only ranking pairs in three metropolitan areas — takes on the joint risk that no offer materializes for both partners simultaneously at any of those pairs. Match rates above 90 percent reflect couples who ranked broadly enough to hedge against that risk.
The single most important strategic feature of the Couples Match — and the one most under-discussed in the literature — is the “No Match” code. Either partner can include “No Match” as a paired combination, meaning that at a given rank, one partner accepts being unmatched in order to allow the other partner to match at the paired program.
How the No Match option works in practice. Suppose Partner A is applying to a competitive specialty (e.g., Diagnostic Radiology) and Partner B is applying to a less competitive specialty (e.g., Family Medicine). At Rank 30 of the paired list, the couple ranks Partner A at Top-Choice Radiology Program / Partner B at “No Match.” If that pair triggers, Partner A matches at the desired program; Partner B accepts being unmatched at that rank and proceeds to SOAP.
In the 2024 Match, 87 couples matched using this pattern — representing about 7 percent of all couples that year. The mechanism is rare but powerful: it lets couples preserve the specialty/program priority of one partner without forcing the other partner to settle for a worse fit.
For IMG couples, the No Match option carries particular weight. Because IMGs face structural filters that USMD applicants do not, the program pool for which both IMG partners are competitive may not include the strongest individual fit for one partner. The No Match option allows the couple to strategically prioritize one partner’s top fit at certain ranks, while still ranking pair combinations elsewhere on the list to maximize the joint match probability.
The trade-off is real: a partner who accepts “No Match” at certain ranks must be prepared to enter SOAP, where IMG match rates are substantially lower than in the main Match. The decision to use the No Match option should be made deliberately, with explicit conversation between partners about acceptable outcomes.
A peer-reviewed analysis published in the Postgraduate Medical Journal in November 2025 provides the most rigorous data to date on Couples Match outcomes specifically within Internal Medicine. Diaz and colleagues surveyed standard-Match and Couples-Match Internal Medicine residents and found three statistically significant differences between the two groups:
These findings reframe the headline 93 percent match rate. The Couples Match is not simply a tool that lets couples match together at no cost — it is a tool that lets couples match together at the cost of placing lower on their rank list, accepting more compromise, and tolerating higher process stress. For IMGs already navigating SUVY filters and visa sponsorship constraints, this is meaningful: the Couples Match is a strategic tool, not a free upgrade, and its effectiveness depends on disciplined planning.
Citation: Diaz A, Singh A, Bujnowski D, Farnan J. NRMP couples match outcomes for internal medicine residents. Postgrad Med J. 2025 Nov 20:qgaf198. doi: 10.1093/postmj/qgaf198.
The NRMP algorithm does not require programs to be aware of an applicant’s Couples Match status. The algorithm operates independently of program awareness. However, applicants may choose to disclose Couples Match status at several points in the application timeline, and disclosure decisions have strategic consequences worth considering.
Common disclosure points include:
For IMG couples, early disclosure can produce two effects worth weighing carefully. On one hand, programs that know two applicants are coupled may consider the joint logistics of bringing both partners into the program (or to nearby programs in the same metropolitan area) and may advocate within their institution to make a match work. On the other hand, programs that do not sponsor visas, or that have limited capacity for joint hires, may filter coupled applicants out earlier rather than later.
The strategic question is whether disclosure helps or hurts at a specific program. For programs already in your viable pool with strong IMG track records and dual sponsorship, disclosure typically helps. For programs at the edge of the viable pool, disclosure can serve as a structural filter. The decision should be program-specific, not blanket.
Pulling the analysis together, the Couples Match for IMGs reduces to four practical principles:
Plan from Stage 1, not Stage 6. The viable program pool for an IMG couple is determined long before the rank order list is submitted. Where each partner applies, where each partner interviews, and the geographic overlap of those two interview lists set the ceiling on what the Couples Match algorithm can do. Couples that begin coordinating in February, when the ROL is due, have already locked in most of the constraint.
Identify the joint viable pool empirically, not optimistically. Map both partners’ SUVY profiles against the same program list. Programs that clear for both go on the joint shortlist. Programs that clear for only one partner must either be dropped or paired with the No Match option for the non-clearing partner.
Treat visa coordination as a Stage 1 binary filter, not a soft preference. If both partners require sponsorship, the joint pool narrows substantially. If sponsorship requirements differ between partners, the joint pool narrows further. Programs that sponsor both visa types must be identified early in application planning — not discovered during interview scheduling.
Use the No Match option deliberately. For couples with asymmetric specialty competitiveness or asymmetric SUVY profiles, the No Match option is not a fallback — it is a strategic mechanism for preserving one partner’s priority while maximizing the joint match probability. Discuss it explicitly. Decide together. Document the agreement before the ROL is certified.
IMGPrep advises IMG couples on coordinated SUVY mapping, joint program list construction, visa-aware program selection, and paired ROL strategy. Two decades of experience with the structural filters that determine IMG match outcomes — applied to your specific paired application.
No. The NRMP states that “any two applicants can choose to participate in an NRMP Match as a couple.” Spouses, romantic partners, friends, classmates, colleagues, and family members are all eligible. The only requirement is that both individuals be eligible for placement in a U.S.-based residency program through the NRMP.
Not directly — couples match rates have remained above 90 percent every year since 1984, including for non-USMD couples. However, peer-reviewed evidence shows Couples Match applicants are significantly less likely to match at one of their top three ranked programs and report significantly higher process stress. For IMGs, who already face structural filters at the binary and quantitative stages of the Match Funnel, the Couples Match adds a layer of constraint that requires deliberate strategic planning to navigate.
No. The Couples Match is administered through the NRMP’s algorithm and does not require any special action from residency programs. Programs evaluate each applicant individually based on qualifications. Programs may be aware of an applicant’s couples status if it is disclosed on the ERAS application or during the interview process, but coupling does not change the program’s evaluation framework.
There is no fixed answer, but data and structural analysis suggest IMG couples should rank as broadly as the joint viable pool supports. A couple ranking only 10–15 paired combinations takes on substantial joint match risk. Couples with successful matches typically rank 30–80 paired combinations, with the highest-priority pairs at the top of the list and broader, more flexible pairs lower down. The base NRMP fee covers up to 20 unique program codes per partner; additional codes are $30 each.
If no paired combination on the rank order list produces a match for both partners, the algorithm does not fall back to matching either partner individually. Both partners go unmatched and proceed to SOAP. This is the structural risk of coupling, and it is why ranking broadly — including via the No Match option for asymmetric pairs — is a critical strategic decision.
Yes. Either partner can initiate uncoupling at any time up to the rank order list certification deadline. The couple registration fee is non-refundable. After certification, the rank order lists are locked and the couple must proceed as paired through the algorithm.
IMGPrep is not associated with the NRMP®, the MATCH®, the ECFMG®, or the AAMC.