How to Find a Biological Relative for a Bone Marrow Match
The national system for finding a marrow or blood stem cell donor was built around volunteers who are strangers to the patient – that is its purpose in the statute, and it is the route your transplant team will be working. A search for a biological relative is the narrow exception – and this page is about the handful of situations where it genuinely earns its place, what the sibling odds actually mean, and who is supposed to be driving all of it.
Thirty seconds, if that is all you have
Tell the transplant center before you do anything else. Say that a biological relative exists whom you cannot currently reach, and ask what they want you to do with a name and address if you get one. The center, not you, submits searches and arranges tissue typing. Understand what the sibling number means. Each full sibling has about a one in four chance of being a full HLA match. That is per sibling, so the value of finding one more is real but bounded. Do not stop the registry search while you look. The national program exists precisely because most patients have no matched relative, and pausing it costs time you cannot buy back. Nobody you find is obliged to be tested. Donation is voluntary at every stage, and a relative who says no has done nothing wrong.
Watch: where a family search actually fits
A short walkthrough of the registry route, the sibling exception, and the point at which a records search stops being a distraction and starts being useful.
Watch the overview
The national program was built around strangers, on purpose
Read the statute and the whole shape of this problem becomes clear in one sentence.
The federal law that creates the national donor program is 42 U.S.C. 274k. Subsection (a) directs the Secretary of Health and Human Services, acting through the Administrator of the Health Resources and Services Administration, to establish and maintain by contract a C.W. Bill Young Cell Transplantation Program, successor to the National Bone Marrow Donor Registry, “that has the purpose of increasing the number of transplants for recipients suitably matched to biologically unrelated donors of bone marrow and cord blood.”
Biologically unrelated. Those two words are the reason this page opens the way it does. The system Congress funded is not a family-finding service that gave up and went looking elsewhere. Its stated purpose, in the text of the law, is the volunteer stranger. The registry side of that program is operated under federal contract by the National Marrow Donor Program, whose donor-facing identity most people know as Be The Match, and it is very large: a 2025 paper in the journal Transfusion reports the World Marrow Donor Association’s count of 44.1 million potential stem cell donors registered worldwide by the end of 2023.
Against that, a family is a very small pool. The transplant literature commonly puts the share of patients who have an available matched sibling donor at roughly thirty percent – the figure appears in this form in both a 2025 comparative outcomes study and the 2025 Transfusion paper above, each citing earlier work rather than measuring it fresh. Turn it around and the point lands: something like two thirds to seven tenths of patients were never going to find their donor inside the family, whatever they did.
So why does this page exist at all? Because a national registry, however large, does not deliver evenly, and because for a specific and identifiable group of patients an unfound relative is worth real effort. The statute itself concedes the first half. Subsection (e)(1) requires the program to identify populations underrepresented among enrolled donors and to prioritize recruiting them “in order to enable a member of such a population, to the extent practicable, to have a probability of finding a suitable unrelated donor that is comparable to the probability that an individual who is not a member of an underrepresented population would have” – and names racial and ethnic minority groups, including persons of mixed ancestry, as such populations. Congress would not have written that clause if the odds were flat.
Recent data puts numbers on the gap. A 2024 study in Blood Advances by Fingrut and colleagues at Memorial Sloan Kettering looked at 3,843 unrelated donors requested for 455 consecutive adult patients between 2020 and 2022. Of the donors requested for confirmatory typing alone, 1,894 of 3,529, or 54 percent, were actually available; domestic availability was 45 percent; and by patient ancestry, availability for confirmatory typing ran 150 of 458 for patients of African ancestry, 120 of 258 for non-Black Hispanic patients and 119 of 270 for Asian patients – 33, 47 and 44 percent respectively. The authors’ own summary is that the functional size of the unrelated donor pool is much smaller than appreciated.
Read that carefully, because it is the most misquoted kind of number on this subject. Those percentages describe how many of the donors a transplant center asked for turned out to be reachable and willing to proceed, at one center, over one period. They are not the chance that a patient finds a donor, and anyone who repeats them that way has changed their meaning. What they support is a narrower and still important claim: the registry route is slower and less certain for some patients than the headline registry size suggests, and for those patients a biological relative is proportionally more worth finding.
The sibling number, and what it does not say
One in four is correct and almost always misunderstood. The other two branches of the same coin toss matter just as much.
You inherit one set of HLA genes from each biological parent, and that set travels together as a block called a haplotype. Two full siblings are therefore drawing from the same four parental haplotypes, which produces three possible outcomes and not one. As a 2025 review in Trends in Cancer states it, because HLA haplotypes are inherited from each parent through Mendelian inheritance, siblings have a 25 percent chance of being HLA identical at the genotypic level, a 50 percent chance of sharing one haplotype, and a 25 percent chance that they share no HLA haplotypes at all. A 2015 paper in the Journal of Immunology Research puts the first branch the same way, noting that identical HLA between two people is otherwise very unlikely.
Three things follow that consumer pages routinely skip.
First, the 25 percent is per sibling, not per family. It is the chance that one named brother or sister is a full match, and it resets for the next one. If each full sibling is an independent draw, then the chance that at least one of them is HLA-identical is one minus three quarters raised to the number of siblings – roughly 25 percent with one, 44 percent with two, 58 percent with three, 68 percent with four. That is arithmetic following from the figure above, not a finding from a study, and it assumes full siblings and independent inheritance. But it is why an unknown or unreachable sibling is not a rounding error. Finding a third sibling when you have two moves the family’s odds by about fourteen points.
Second, the middle branch is not a failure. A sibling who shares exactly one haplotype is a half match, or haploidentical, and half-matched relatives are now transplanted routinely. Their standing relative to a full match is genuinely contested in the literature rather than settled: a 2018 editorial in Haematologica, reviewing a European registry comparison, concluded that a matched sibling should be selected where one is available, with a haploidentical relative as the suitable alternative when one is not. Which of those two your team reaches for is a clinical judgment about your specific disease and risk, made by physicians. It is emphatically not a question this page is qualified to answer, and you should be suspicious of any non-clinical page that tries.
Third, the third branch is real. One sibling in four shares no haplotype with you and is neither a full match nor a half match. A family can be typed in full and come back with nothing, and that outcome is ordinary rather than unlucky.
All of which is the honest case for a search. The registry cannot enroll a person it does not know exists, and it will never contain a sibling who was placed for adoption in 1974, or who left the family after a divorce and has not been heard from since, or whose existence surfaced last month in a DNA match list. That is the one thing a records search can add that a 44 million person registry structurally cannot: a specific person, with a specific claim to a quarter of your haplotypes, who is currently not in anybody’s system. If that describes your situation, our guidance on tracing a sibling you have lost contact with covers the record trail in more detail than belongs here.
Half-siblings, donor conception, and the parent haplotype
Sharing one parent changes the genetics in a way that is easy to state and often explained badly.
A great many people meet this subject for the first time in a distressing week, having just learned something about their family they did not previously know. Consumer DNA testing has made that extremely common, and a patient who discovers half-siblings, or discovers that they were donor-conceived, quite reasonably wants to know what those relationships are worth here.
Start from the inheritance, and the answer is clean. Each of your two haplotypes came from one biological parent. A half-sibling shares only one parent with you, so there is only one side on which the two of you could have drawn the same block. The most a half-sibling can share with you by descent is that single haplotype, and whether they did is a coin toss. The full-sibling 25 percent has no counterpart in a half-sibling, because there is no second shared parent for a second matching haplotype to come from. What a half-sibling can realistically be is a half match, at roughly the same odds as a full sibling on that one shared side.
Biological parents and biological children are a different case, and a more certain one. Because every child inherits exactly one haplotype from each parent, a parent and child always share exactly one – they are obligate half matches, with no coin toss involved. A patient’s mother, father and children are therefore the most predictable haploidentical donors in any family, subject entirely to the medical criteria that only a transplant team can apply. Age, health and the physician’s judgment decide whether any of them is actually suitable; genetics only decides that the half match is there.
For donor-conceived patients this reframes the search. The half-siblings surfaced by a testing service are not candidates for a full match, and it is kinder to know that up front than to build hope on it. They are candidates for a half match, and there may be a considerable number of them. The biological donor, if identifiable, is an obligate half match in the same way any biological parent is. None of that changes the order of operations – registry first, family typing arranged by the transplant center – but it does tell you which names on a match list are worth the effort of tracing. We deal with the practical end of this in our guidance on locating a half-sibling through the shared parent and, where the route runs through a clinic or a registry rather than a family, on tracing a biological parent after donor conception.
One caution on the testing services themselves. A centimorgan estimate and a predicted relationship label are not tissue typing and cannot be substituted for it. HLA typing is a separate laboratory test, ordered through the transplant center, on a sample the person gives knowingly and for that purpose. A consumer DNA report tells you that a relationship probably exists. It tells you nothing usable about whether that person is a match.
The transplant center drives this, and there is a named person
Weeks get lost here. Families try to work the registry themselves, and the registry is not built to be worked that way.
The donor search is submitted by the transplant center, not by the patient. A 2022 report from the National Marrow Donor Program published in Frontiers in Immunology measures its own workload in exactly those terms, counting the preliminary and formal donor searches submitted from transplant centers to the program. Note that there are two stages with different meanings: a preliminary search is a look at what the registry holds against your HLA type, and a formal search is the point at which specific donors are pursued, contacted and asked for confirmatory samples. Knowing which stage you are at is one of the most useful questions you can ask, and almost nobody thinks to ask it.
The person to ask is the search coordinator. That is a real staffed role inside a transplant program rather than an informal title. A 2026 study in Transplantation and Cellular Therapy examining donor source selection across fifty United States transplant centers drew its interviewees from physicians, advanced practice providers, medical directors, search coordinators and HLA laboratory staff, and international recommendations on establishing a transplant program list access to a donor search coordinator among the requirements for running allogeneic transplants at all. If you have been given a name and a direct number for anyone in the program, the coordinator is the one whose job includes your donor search specifically.
There is also a statutory advocate, and hardly anyone uses it. Section 274k(h) requires the Secretary to establish and maintain an Office of Patient Advocacy, headed by a director and staffed by people with expertise in the therapy, operating a patient advocacy system that the statute expressly requires to be separate from the mechanisms for donor advocacy. It serves patients for whom the program is conducting, or has been requested to conduct, a search. Its listed duties are concrete: periodically inform the patient about suitably matched donors and the progress of the search, inform the patient if the search has been interrupted or discontinued, identify and resolve problems in the search where practicable, and make available a comparison of transplant centers on the search costs charged to patients before transplantation and on post-transplant outcomes. Separately, subsection (d)(1)(I) requires the program to collect, analyze and publish data on how far patients get through the search process, how many never complete it, and why.
That is an accountability structure that exists whether or not anyone tells you about it, and the clause about being told if a search is interrupted is the one worth remembering. Silence from a search is not the same as a search still running.
What we will not do is put a week count on any of this. Published summaries of donor search timelines vary widely, and we could not verify a specific range from a source we read at first hand, so we are not going to invent one on a page people read while frightened. What is verifiable is the sequence – preliminary search, formal search, confirmatory typing, donor workup – and the fact that the center running it is the only party who can tell you where in that sequence you are and what date they are working toward. Ask them, and ask again if the answer stops arriving.
What each route can supply, and what it cannot
Ordered by how the system actually uses them, not by how promising they sound from outside.
| Route | What it can actually supply | What it cannot |
|---|---|---|
| The national registry | Volunteer unrelated donors and cord blood units, searchable by the center against your HLA type, plus a statutory patient advocacy office | Cannot contain a relative it has never enrolled; availability of requested donors is uneven and lower for several ancestry groups |
| A full biological sibling | The only relative who can be HLA-identical by descent, at about one chance in four each | Roughly seven in ten patients have no available matched sibling at all; typing must be arranged through the center |
| A parent or a biological child | An obligate half match every time, with no genetic uncertainty at all | Never a full match by descent; medical suitability is a separate question the transplant team decides |
| A half-sibling | A possible half match on the shared parent’s side, roughly a coin toss | Cannot be HLA-identical by descent, because only one parent is shared |
| Public-records research | An address and contact route for a specific relative nobody can currently reachNarrow role | No access to HLA data, medical records, registry files or transplant center systems; no ability to make anyone agree to anything |
| Consumer DNA testing services | Evidence that a biological relationship probably exists, and sometimes a name | Not tissue typing, and not a substitute for it under any circumstances |
Our own row is deliberately the modest one. A firm that does this for a living should be plain that in a marrow case it holds no clinical position whatsoever: we cannot type anyone, cannot read a match report, cannot see a registry record and would refuse a request to try. The entire contribution is upstream of the medicine – turning a name, a decades-old address and a fragment of family memory into a current, documented address that a family or a transplant center’s staff can act on. Where that is the missing piece, our skip tracing and public-records research is the right tool. Where it is not, it is an expensive distraction from the search the center is already running.
Urgency gets more scrutiny here, not less
This has to be said out loud on a page like this one, and it belongs in the argument rather than in a footer.
A medical emergency involving a child, a sibling or a parent is the most compelling reason a person can give for wanting somebody found. That is exactly why it is the reason most often borrowed. Anyone who takes locate requests for a living learns quickly that the most urgent, most sympathetic and most time-pressured story in the inbox is not automatically the truest one, and treating it as self-evidently true is how a research firm ends up handing a stalker an address.
So we do the opposite of what the pressure suggests. On this subject the intake questions get longer, not shorter. We ask who the patient is and what the requester’s relationship to them is. We ask which transplant center is involved and whether its team knows this search is happening. We ask whether the person being sought has ever declined contact before. None of that is bureaucracy for its own sake, and none of it is negotiable because a case is moving fast.
Where a request touches safety, we decline it. If there is a protective order, a stalking allegation, a contested custody matter, an address confidentiality program enrollment, or a person who has already told the requester they do not want contact, we do not take the file – and we do not take it regardless of what medical circumstance is attached. Someone who is hard to find is sometimes deliberately hard to find, and the reasons are frequently good ones. In those cases the right routes are the transplant center’s own social work team, a victim services advocate, or the court that issued the order, and we will say so rather than quietly proceed. People live where they live, and a person who moved away from someone is not a fugitive from them.
We also do not tell anyone why we are looking. A diagnosis belongs to the patient. Nobody we speak to in the course of research learns that there is a patient, what the illness is, or that a transplant is being contemplated – not a former neighbor, not a landlord, not a relative on an adjacent branch of the family. If the search succeeds, the news is delivered by the family or by clinical staff who are trained to deliver it, not by a stranger from a records firm leaving a voicemail. That rule holds even when disclosing would obviously make the research easier, which it often would.
And finding somebody is not recruiting them. A located relative is under no obligation to be swabbed, typed, evaluated, or to explain a decision either way. Donation is voluntary at every stage and remains so right up to the end. This is not merely our house view: the field’s own ethics framework builds it in. The American Academy of Pediatrics Committee on Bioethics criteria for when a minor may ethically serve as a stem cell donor open with the condition that there is no medically equivalent adult relative who is willing and able to donate, and the same guidance has the independent donor advocate involved from the very start, including in the decision about whether to type a person at all, so that family or sibling donors with medical or psychological reasons not to donate are not typed in the first place. Pressure applied to a relative is both wrong and, on the evidence of how these programs are designed, counterproductive.
Six shapes this arrives in, and where each one starts
Find the one that matches your situation. The first move is different in each.
You were adopted and have never met your biological family
Ask the transplant center first: many programs have staff who handle exactly this, and some states allow a medical-necessity route to adoption records that is faster than a records search. Start a parallel search only once you know what the center can already do for you.
A sibling exists and the family lost touch decades ago
This is the clearest case for records work, because the person is not hiding and there is a documented trail. Bring the last known address, the approximate year of the last contact, any spelling of the name they used, and the names of the people who knew them.
A DNA test just produced half-siblings you did not know about
Useful, but calibrate. Half-siblings cannot be HLA-identical to you by descent; they are candidates for a half match on the shared parent’s side. Give the list to the transplant center before you spend weeks tracing everyone on it.
You are donor-conceived and want to reach the donor
A biological parent is an obligate half match, so the genetics are on your side even though a full match is off the table. The route usually runs through a clinic, a donor registry or a testing service rather than through public records alone, and clinic-era record retention is the usual obstacle.
You found them and the answer was no
That is a complete and legitimate answer, and it does not need a reason attached to it. Do not send a second person to ask again. Tell the transplant center, which will keep working the registry and the half-matched relatives who did consent.
The relative left this family, and left for a reason
This is the request we decline, and we decline it whatever medical urgency is attached. Protective orders, stalking histories, contested custody and a prior refusal of contact all stop the file at intake. Those questions belong with the court, an advocate, or the center’s social work team.
How a locate like this runs on our side
Four steps, and the first one frequently ends with us telling you not to hire us.
Confirm the clinical route is already moving
The opening conversation is a triage and it costs nothing. If the registry search has not been started, if the center has not been told this relative exists, or if the program’s own staff can obtain the records faster, we say so and send you back there first. Taking a fee for work a transplant program would do better is not a service.
Establish who is asking and on what basis
We record the requester, the relationship, the lawful purpose the research serves and the safety questions set out above. A file that cannot clear that step does not open, and no amount of urgency changes it. Nothing proceeds on a first name and a hope.
Rebuild the trail from public and licensed records
Address history, property and court filings, licensing footprints, phone attribution and the relatives and associates recorded around the person. Long-separated families throw off a lot of near-misses, so identity is confirmed against several independent records before a name is treated as the right one.
Hand over a documented address and stop
You get the current address, the date it was established, the record behind each element and an explicit list of what could not be confirmed. We do not make the approach and we do not disclose why anyone is looking. Who makes contact, and how, is decided by you and the clinical team.
What we are, and the limits that come with it
Stated before you engage us, because on a subject this consequential the limits are most of the useful information.
We are a skip tracing and public-records research firm. Nobody on this team holds a private investigator’s license, we do not run surveillance, and we hold no credential that opens a door the public cannot lawfully open. Everything we work from is either a public record or a data source we are licensed to query under a purpose the law permits, and every file records that lawful purpose in writing before a single search is run. Where a purpose cannot be stated, there is no file.
We do not obtain information by pretexting. We do not call anyone pretending to be a hospital, a registry, a relative, an attorney or a government office, and we do not impersonate anybody to anybody for any reason. We do not obtain the contents of private financial accounts – no balances, no statements, no transaction histories. And on this subject specifically: we have no access to HLA typing data, medical records, transplant center systems, registry files or donor identities, we cannot enroll anyone on a registry or accelerate a search, and we would decline a request to attempt any of it. Anyone telling a family otherwise during a crisis is either mistaken or selling something worth refusing.
We are not a consumer reporting agency and nothing we produce is a consumer report under the Fair Credit Reporting Act. Our results may not be used for decisions about credit, insurance underwriting, employment or the screening of a job applicant, tenancy or housing, or eligibility for any other benefit the FCRA covers. A family locate is not one of those uses, but people sometimes arrive on a page like this one carrying a different need, and if yours is a covered use the correct provider is an FCRA-compliant consumer reporting agency, which we will tell you instead of taking the work.
Nothing here is medical advice, and we are careful not to look like we are giving any. Whether a particular relative is a suitable donor, whether a half match or an unrelated donor is the better choice for a specific disease, and what any timeline should be are questions for physicians who have seen the case. The figures on this page are cited so you can read the sources yourself, and we have deliberately published no number we could not verify at first hand – which is why you will not find a donor search timeline here. This page is general information and not legal advice; access rules for adoption, vital and clinic records vary by state and change, so confirm anything that matters with the agency or with counsel. The safety decline set out above is absolute and is not reconsidered on a hard day.
One last practical note, since it comes up constantly in family searches of this kind: locating somebody and hearing back from them are separate problems, and the second is often the harder one. Our guidance on what to do when a match does not reply is written for that gap, and it applies here with the extra caution that a single unanswered message is not permission to escalate.
Who brings us a case like this
Nearly always somebody with a specific name in mind and no way to reach it – not a general appeal for help.
Patients with a sibling nobody has heard from
One name, one last known city, and thirty years of silence in between
Adoptees facing a treatment decision
People who need a biological family they have never met, on a clinical clock
Donor-conceived adults
Match lists full of half-siblings and no reliable way to reach any of them
Parents searching on a child’s behalf
Looking for an estranged aunt, uncle or half-sibling of a pediatric patient
Hospital social workers and case managers
Staff who have exhausted the contact information in the chart
Families building a medical history
Relatives assembling what the biological family has been diagnosed with
What links them is that the missing piece is a current address and nothing else. The relationship is known, the name is known, and the reason is one the person being sought would very likely accept – which is precisely why the safety questions still get asked, because that description also fits a request from somebody who should not be given the address. Where the goal is a broader picture of what runs in a biological family rather than a specific donor, our guide to tracing biological relatives for a family medical history is the better starting point. The solid-organ version of this question – finding a relative who might donate a kidney or a piece of a liver – works quite differently, and nothing on this page transfers to it: there is no equivalent national registry for living organ donation, and the evaluation the relative would face is surgical.
What we undertake on a search like this
We will tell you before you engage us if the transplant center or a free channel is the better route, and we will send you there when it is. Every address we return arrives with the record it came from, the date it was established and a written list of what we could not confirm. We will not tell anyone why we are looking, we will not approach the person on your behalf, and if the trail runs cold we will say so plainly rather than hand you a plausible guess at a moment when a guess could cost real time.
Questions patients and families actually ask
Does a bone marrow donor have to be a family member?
No. The federal program at 42 U.S.C. 274k was created with the express purpose of increasing transplants for recipients suitably matched to biologically unrelated donors, and the transplant literature commonly puts the share of patients who have an available matched sibling donor at only about thirty percent. A family search is worth doing when a specific biological relative exists and cannot be reached, not as a substitute for the registry search your transplant center is already running.
What is the chance that a sibling is a match?
Each full sibling has about a 25 percent chance of being HLA-identical to you, a 50 percent chance of sharing one haplotype instead, which makes them a half match, and a 25 percent chance of sharing neither. Those figures come from Mendelian inheritance of HLA haplotypes and are stated in that form in a 2025 review in Trends in Cancer. The critical point is that the 25 percent applies to each sibling separately, so a family with four siblings has considerably better odds than a family with one.
What does HLA matching actually mean?
HLA proteins are inherited as blocks called haplotypes, one from each biological parent, and matching compares those markers between patient and donor. Because the blocks travel together, relatives are far more likely to match than strangers, and full siblings are the only relatives who can be identical to you by descent. HLA typing is a specific laboratory test ordered through the transplant center; a consumer DNA report is not the same thing and cannot stand in for it.
Can a half-sibling be a bone marrow match?
A half-sibling can be a half match but not a full match by descent. You share only one biological parent, so there is only one side on which the same haplotype could have been inherited, and whether it was is roughly a coin toss. There is no second shared parent to supply a second matching haplotype, which is why the full-sibling 25 percent has no equivalent here. Half-matched relatives are transplanted routinely, so this is useful information rather than bad news.
Who starts the donor search, me or the hospital?
The transplant center. Searches are submitted from transplant centers to the national program, in two stages: a preliminary search against the registry, and then a formal search in which specific donors are pursued and asked for confirmatory samples. The staff member whose role covers this is the search coordinator. Ask which stage your search is at, because the answer changes what you should reasonably expect next.
Is a matched sibling always the best donor?
Usually, but it is a clinical judgment and the literature is not unanimous. A 2018 editorial in Haematologica reviewing a European registry comparison concluded that a matched sibling should be selected where one is available, with a haploidentical relative as the suitable alternative when one is not. Donor age, the specific disease and risk category all enter into it. This is a question for the transplant team, and we are not qualified to answer it.
What if the relative I find does not want to be tested?
That is their decision to make and it needs no explanation. Donation is voluntary at every stage. The ethics guidance the field itself uses assumes this: the American Academy of Pediatrics criteria for a minor donating begin with the requirement that no medically equivalent adult relative is willing and able to donate, and the accompanying guidance keeps people who have reasons not to donate from being typed at all. Pressing a relative is wrong, and it is also the least effective thing you can do. Tell the center and let the registry search continue.
How quickly can you find someone, and how long does the search take overall?
We come back within 24 hours with a first read on whether the trail is workable and what the records currently show, and quite often that read is a recommendation to let the transplant center handle it instead. What we will not do is publish a timeline for the medical search itself. Reported figures for donor searches vary widely and we could not verify a specific range at first hand, so the honest answer is that only the center running your search can tell you where it stands and what date they are working toward.
One name, and no way to reach it
Keep the registry search running and keep the transplant center leading it – that is the route that works for most patients, and we will say so even when it means no work for us. If there is a specific biological relative nobody can find, that is the part we can help with. Tell us what you already know and we will give you an honest read on whether the record trail goes anywhere.
Start a family locate
