Couples Match: How It Works and How to Protect Your Pair
What Is the Couples Match?
The NRMP Couples Match is a feature of the Main Residency Match that allows two applicants to link their rank order lists so that program offers are evaluated as pairs rather than independently. When you couple, you are not submitting one joint list—each partner maintains a separate rank order list—but you coordinate those lists so that every entry on your list corresponds to a paired entry on your partner's list. The algorithm then treats each pair as a single unit: either both partners match at their paired programs or neither does, and the system moves down to the next pair.
The practical effect is geographic coordination. You give up some individual optimization in exchange for the probability of landing in the same city. That trade-off is the central tension of every couples match strategy decision you will make.
Who Is Eligible to Couple?
Any two applicants registered in the NRMP Main Residency Match may couple. The NRMP places no restrictions on specialty combination, relationship type, MD/DO/IMG status, or whether both partners are applying to the same specialty or different ones. A US MD applying to internal medicine can couple with an IMG applying to psychiatry. Two partners applying to the same competitive surgical specialty can couple. A DO and an MD can couple.
What does matter operationally: both applicants must be registered in the Main Residency Match for the same match year. Applicants participating in other matches—the San Francisco Match for ophthalmology, the Urology Match, the Neurology Match conducted through SF Match—are outside the NRMP system for those specialties and cannot couple those ranks. If one partner is applying to a specialty that uses a match outside the NRMP, that specialty cannot be included in the couples pairing. Plan accordingly if your specialty combination spans match systems.
How the Couples Algorithm Works
The NRMP uses a deferred acceptance algorithm. For couples, the algorithm works through your paired rank list from the top pair down. At each pair, it asks: can a tentative offer be extended to both programs simultaneously? If both programs have a slot available at that moment in the algorithm's processing, the pair is tentatively accepted. If either program cannot accommodate one of the partners, the entire pair is rejected and the algorithm moves to the next pair down the list.
This has a cascading consequence that solo applicants don't face. Because the algorithm processes all applicants simultaneously—not sequentially—a tentative match can be displaced later if a more preferred applicant claims a slot. When one partner in a tentatively matched pair gets displaced, the algorithm must re-evaluate both partners together, moving them to the next viable pair. This cascade can travel several steps down your list before stabilizing. The algorithm iterates until no further improvement is possible for any applicant.
The takeaway for list-building: pairs near the top of your list must be realistic for both partners simultaneously. A pair where one partner's program is a significant reach and the other's is a near-certain match may cascade further than you expect, because the reach program may not hold the tentative offer while the cascade resolves.
NRMP publishes a detailed description of the couples algorithm in its "Addendum to the NRMP Algorithm" document, available on the NRMP website. Reading the official document before you build your ROL is worth the time.
Does Couples Matching Affect Your Probability of Matching?
NRMP publishes match rate data for coupled versus non-coupled applicants each cycle in its Match results publications. Historically, couples who are both US MD seniors match at rates comparable to their non-coupled peers in most specialties. The match rate difference becomes more meaningful when geographic constraint is severe—when a couple restricts their list to a small number of cities—or when one partner is in a highly competitive specialty with limited programs nationally.
The honest framing: coupling itself does not inherently lower your individual probability of matching if you build your list with sufficient geographic flexibility and apply to enough programs. The constraint you impose on yourself—by restricting geography, by applying to too few programs, or by building a list that has very few viable pairs in the middle tiers—is what lowers probability. The algorithm is not penalizing you for coupling; it is faithfully executing the constraints you set.
See the NRMP's published data pages (Charting Outcomes in the Match, Main Residency Match results) for current-cycle figures. We do not reproduce specific match rate percentages in prose because they shift each cycle.
Choosing the Right Specialty Combination
The risk profile of a couples match is largely determined by the less-forgiving specialty in the pair. When both partners are applying to fields with many programs and broad geographic distribution—general internal medicine, family medicine, psychiatry, pediatrics—the list-building problem is tractable. When one partner is applying to a highly competitive, low-volume specialty (neurological surgery, plastic surgery, orthopedic surgery, dermatology, radiation oncology), the couples match becomes materially harder.
The core constraint: in a competitive specialty, your partner may have interviews at only a limited number of programs, concentrated in academic medical centers in a handful of cities. Your paired list is then anchored to those cities. If the other partner's preferred specialty is also geographically concentrated, viable pairs can become scarce quickly.
Mitigation strategies for asymmetric competitive pairings
- The less competitive partner over-applies. If one partner is in dermatology and can realistically interview at a limited number of programs, the other partner needs enough programs in those same cities to generate sufficient pairs at each tier of the list. This often means the less competitive partner applies to more programs than they would solo—sometimes substantially more.
- Geographic flexibility is the most powerful lever you have. Every additional city you are both willing to consider multiplies the number of viable pairs. Quantify this explicitly before you finalize your application lists.
- Candidly assess before you apply, not after. If one partner is applying to a specialty where the realistic interview yield in your acceptable cities is very low, build your ROL strategy around that constraint from the start, not as an afterthought.
Building Your Shared Rank Order List
Each partner submits their own ROL in the NRMP system. When you couple, you link your lists so that each rank on your list is paired with a specific rank on your partner's list. A pair is entered as: "My rank #1 paired with your rank #3." The result is a set of paired entries that the algorithm evaluates in order.
City-first versus program-first strategies
Couples generally build their lists using one of two organizing logics, or a hybrid.
- City-first: You identify your preferred cities in order, then rank all viable program pairs within each city before moving to the next city. This maximizes geographic coherence. The risk is that you may be ranking a lower-quality program for one partner high on the list because it happens to be in the preferred city, even when a better program for that partner exists in a less preferred city.
- Program-first: Each partner ranks their programs by quality/fit independently, then you construct pairs by matching entries from each list. This maximizes individual program quality but may produce pairs that are geographically incoherent at the top of the list.
Most couples use a tiered hybrid: city-first within preference tiers, program-first within cities. Define your city tiers explicitly and in agreement before you start pairing. Disagreement about city preference that surfaces late—after interview season—is a primary source of couples match ROL conflict.
Handling asymmetric list lengths
Your lists will almost never be the same length. One partner may have interviewed at more programs. The NRMP system accommodates this: you can pair one program on your list with a "null" entry on your partner's list, meaning your partner has no paired entry at that rank. A null pair is not the same as a solo rank—see the next section on solo ranks. A null pair means the pair can only match if your partner matches elsewhere through a different pair. Work through the NRMP's official couples registration guidance to understand how null entries function mechanically before you build your list.
Complementary pair tiers
Structure your paired list in tiers that reflect joint preference, not just individual preference. A useful exercise: independently rank your cities 1–N and your programs within cities 1–M. Then sit down together and build pairs that reflect joint preference at each tier. Pairs where one partner is enthusiastic and the other is reluctant tend to cause cascading regret more than cascading algorithm problems—but they cause both.
How Many Programs Should Each Partner Apply To?
There is no universal number, and any specific figure we publish here will be outdated within a cycle. NRMP and specialty-specific sources (AAMC Charting Outcomes, specialty society data) publish application volume data with match outcome correlation. Use those for current benchmarks.
The couples-specific principle: the less competitive partner in the pair often needs to apply to more programs than their solo equivalent would, specifically to generate enough geographic overlap with their partner's interview pool. This is a real cost—financial and logistical—but it is the primary mechanism for creating viable pairs across multiple cities and tiers.
A practical approach:
- After interview invitations arrive (but before you accept or decline), map where each partner has interviews by city.
- Count how many cities have overlapping interviews for both partners. This is your raw pair-generation pool.
- If that pool is thin—few overlapping cities, or overlap concentrated in only your top choices—consider whether either partner should pursue additional applications through programs with late-opening interview slots, or accept that your list will be short and plan accordingly.
- A short but honest list is better than a long list padded with pairs neither partner would accept. The algorithm ranks you as high as your list goes; it cannot help you beyond it.
Interview Scheduling as a Couple
Interview scheduling is where couples match strategy becomes logistically demanding in a way that solo applicants don't experience.
Coordinating dates in the same city
When both partners have interviews in the same city, attending them on consecutive days (or the same day if programs allow) minimizes travel costs and time. This requires active calendar coordination throughout interview season, which begins before either partner has a complete picture of their interview pool. Build a shared calendar from the first invitation and update it in real time.
Accept that you will not always be able to coordinate perfectly. One partner may need to travel to a city alone while the other attends interviews elsewhere. Don't sacrifice a high-priority program for scheduling convenience.
Whether to disclose couple status during interviews
You are not obligated to disclose that you are participating in the Couples Match during interviews. Programs cannot ask about it, and you have no obligation to volunteer it.
Some applicants choose to disclose, particularly when asking a program coordinator about scheduling flexibility or when they feel the disclosure explains a genuine question about the city. If you disclose, disclose to the coordinator or program director as a logistical note, not as a plea for preference. Programs cannot give preference to coupled applicants and cannot penalize you for coupling.
The practical risk of disclosure is low in most contexts. The practical benefit is also generally low. Make the decision case by case based on whether it serves a specific logistical purpose.
The Solo Rank: Ranking Programs Apart
Each partner may include unpaired, solo ranks at the bottom of their rank order list. A solo rank means: if all of my paired entries have failed, I am willing to match at this program alone, without my partner. The algorithm will attempt to match you there independently only after all paired entries have been exhausted.
Solo ranks are a genuine safety net, and whether to include them is one of the most important strategic decisions couples make.
Arguments for including solo ranks:
- They protect against an outcome where both partners go unmatched because no pair was viable, even though one or both could have matched solo.
- They are particularly valuable when one partner is in a highly competitive specialty—if that partner cannot match anywhere as part of a pair, a solo rank at a program they would accept gives them a path to not going unmatched while the other partner enters SOAP.
Arguments against including solo ranks, or for placing them very low:
- Including a solo rank you would not genuinely accept creates a commitment problem. If you match there, you are contractually obligated to attend.
- If your paired list is well-constructed with sufficient geographic flexibility, the need for solo ranks is reduced.
The decision should be made jointly and honestly: are there programs either partner would genuinely accept as a solo match if the alternative is going unmatched? If yes, rank them. If the honest answer is no, do not include them as a formality—the obligation is real.
SOAP for Couples
The Supplemental Offer and Acceptance Program (SOAP) opens for applicants who did not match. For couples, SOAP eligibility and mechanics work as follows:
- Each partner's SOAP eligibility is determined individually, based on their own match outcome. An applicant who did not match is eligible for SOAP; an applicant who matched is not.
- If both partners did not match, both are eligible for SOAP. However, SOAP does not have a couples coordination mechanism. You are each competing for available positions independently.
- SOAP positions are limited, distributed across programs with unfilled slots, and the process moves very quickly. Partners who both enter SOAP may not be able to coordinate geographically during SOAP—programs are not required to accommodate couples in SOAP, and the timeline does not allow extended deliberation.
- If one partner matched and the other did not, the unmatched partner enters SOAP alone while the matched partner fulfills their match commitment.
This asymmetric outcome—one matched, one not—is the scenario couples find hardest to plan for and hardest to experience. It is not rare enough to dismiss. Build your paired list with enough viable pairs that this probability is low, but discuss explicitly what you will do if it occurs. The SOAP process moves on a compressed timeline; decisions made under that pressure without prior agreement tend to be harder.
NRMP publishes SOAP rules and eligibility criteria on its website. Verify current SOAP mechanics directly with NRMP for your application year, as procedural details can change.
Timeline and NRMP Registration Steps
The following describes the process in general terms. Specific deadlines change each cycle; see the current season timeline on the NRMP website and our data pages.
- Register in the Match: Both partners must register individually with the NRMP for the Main Residency Match.
- Link as a couple: The NRMP provides a couples registration function within R3 (the NRMP applicant portal). One partner initiates the coupling request; the other must confirm it. Both must be registered before linking is possible. Complete this step early in the cycle—do not leave it to the ROL submission period.
- Build paired ROLs: Work through your paired list structure together before entering it into the system. Entering it correctly requires that you understand which rank on your list pairs with which rank on your partner's list.
- ROL submission deadline: The NRMP sets a specific deadline each cycle by which your rank order list must be certified. Late or uncertified lists are not processed. Both partners must certify independently.
- Common errors to avoid: Failing to confirm the coupling link (one partner initiates, the other never confirms); certifying your ROL before your partner has finalized theirs and discovering you need to make changes; entering pairs in the wrong order due to miscommunication about which rank numbers correspond.
Work through the NRMP's official couples documentation in R3 before you begin entering your list. The interface is functional but requires attention to how pair entries are structured.
Couples Match for IMGs and DOs
IMG and DO applicants are eligible to participate in the Couples Match under the same NRMP rules as US MD applicants. There are, however, practical considerations specific to each group.
International medical graduates
IMGs must meet ECFMG certification requirements to participate in the Main Residency Match. Coupling does not change those requirements. An IMG applicant who is not ECFMG-certified cannot register for the Match and therefore cannot couple.
IMG applicants often face a more restricted interview pool than US MD applicants in the same specialty, which can limit the number of viable geographic pairs. This is not a function of the couples mechanism—it is a function of interview volume—but it matters more in the couples context because each city with interviews represents potential pairs. An IMG applicant with a smaller interview pool has fewer cities to work with, which tightens the list-building problem.
When one partner is an IMG and the other is a US MD, their interview pools may be geographically skewed in different ways. Map the overlap explicitly after interview invitations arrive.
Verify current requirements directly with ECFMG/Intealth and official sources for your application year.
DO applicants
Following the integration of AOA and ACGME accreditation systems, DO applicants participate in the NRMP Main Residency Match on the same basis as MD applicants for ACGME-accredited programs. DO applicants may couple using the same mechanism. There is no longer a separate AOA Match for most specialties, though applicants should verify the current accreditation status of any program they are considering.
IMG + US MD couples: asymmetric risk
When one partner is a US MD and the other is an IMG, the partners often carry different baseline match probabilities in their respective specialties, independent of the couples mechanism. This asymmetry is worth quantifying honestly before you build your strategy. The partner with the more constrained interview pool—regardless of which partner that is—anchors the geographic constraint for the pair. Plan your application volume and geographic flexibility around that constraint.
Real Questions to Ask Yourselves Before You Couple
Coupling is not automatically the right choice for every pair of applicants who are in a relationship. It is a strategic decision with real consequences. Work through these questions together before you register as a couple.
How geographically flexible are you, honestly?
The word "flexible" is easy to say in October and harder to mean in February when the city on your list is not where either of you imagined. Be specific: which cities would you both genuinely accept? Which would one of you accept but the other would not? Build your list around the honest intersection, not the aspirational one.
What is your plan if one partner doesn't match?
SOAP is fast, geographically unpredictable, and does not have a couples coordination mechanism. If one partner matches and the other does not, you may face a year apart or a significant geographic compromise under time pressure. Discuss this before Match Day, not during SOAP week.
What is your plan if neither partner matches?
Both entering SOAP simultaneously, with independent searches for open positions, is possible. Discuss whether you would prioritize matching in the same city even if it means one partner takes a less preferred program, or whether each partner will pursue the best available option independently.
Should you decouple if circumstances change?
The NRMP allows couples to decouple before the ROL deadline. If your relationship circumstances change, or if one partner's interview pool is so limited that coupling materially increases the risk of going unmatched for both, decoupling is a legitimate option. This is not a failure of the relationship or the strategy—it is a rational response to changed information. The decision to decouple should be made well before the ROL deadline, with enough time to rebuild individual rank lists.
Are you coupling because it serves your match strategy, or because it feels like the right thing to do?
These motivations are not always aligned. Coupling when one partner is in a highly competitive specialty with a very small interview pool, in a field where geographic concentration is extreme, and the other partner has strong solo match prospects may lower both partners' match probability without a commensurate geographic benefit. Assess this candidly. A year apart, with a clear plan to reunite, may produce better long-term outcomes than both going unmatched or one partner accepting a program they would not have ranked solo.
There is no right answer to this question for every couple. There is only the answer that reflects your actual preferences, your actual risk tolerance, and your actual flexibility—assessed honestly before you commit to a list.