What Are the Odds of Being a Bone Marrow Match?

The odds of being a bone marrow match depend heavily on who you are comparing. Between any two unrelated people picked at random, the chance of a match is less than 1 in 10,000. Between siblings, it jumps to about 25 percent per sibling, though the real-world probability of having at least one matched sibling in your family ranges from roughly 13 to 51 percent depending on your age, race, and family size. For patients searching an unrelated donor registry, the likelihood of finding a suitable match ranges from as high as 75 percent for white patients of European descent to as low as 16 percent for Black patients of South or Central American descent. These gaps are not arbitrary; they trace back to the staggering genetic diversity of the immune system markers that must align for a transplant to work.

What “Matching” Actually Means

When doctors talk about a bone marrow match, they are talking about human leukocyte antigens, or HLA. These are proteins on the surface of your cells that your immune system uses to distinguish “self” from “foreign.” If a donor’s HLA markers are too different from the patient’s, the transplanted cells may attack the patient’s body, a potentially fatal reaction called graft-versus-host disease. The key HLA genes that transplant teams focus on are HLA-A, HLA-B, HLA-C, and HLA-DRB1. Some centers also evaluate HLA-DQB1 and HLA-DPB1, bringing the total to as many as twelve markers (two copies of each gene, one from each parent) that need to line up.

A “10/10 match” means both copies of five HLA genes are identical between donor and patient. An “8/8 match” looks at four genes. The more markers that align, the better the transplant tends to go. High-resolution mismatches, where the proteins are slightly different at a detailed molecular level, carry less risk than low-resolution mismatches, where the proteins differ in broader, more obvious ways.¹ This is why modern tissue typing uses DNA-based methods rather than the older serological tests that could only detect broad categories of HLA types.²

Sibling and Family Match Odds

Each biological sibling inherits one set of HLA genes from each parent, bundled into two “haplotypes.” Because there are only four possible haplotype combinations between two parents, each sibling has a one-in-four chance of being a perfect HLA match with any other sibling. That 25 percent figure per sibling is a starting point, but it does not tell you the likelihood that at least one sibling in your family is a match. With two siblings, the odds are about 44 percent. With three, about 58 percent. The more siblings, the better, but modern families tend to be smaller than in past decades.

The commonly cited estimate in transplant medicine has long been that about 30 percent of patients will find a matched sibling. But a study using U.S. birth data and statistical modeling found that this figure masks enormous variation. Depending on patient age and race or ethnicity, the actual probability of having an HLA-identical sibling ranged from 13 percent to 51 percent.³ Older patients tend to have better odds simply because they were born in an era of larger families. Younger patients, especially those from smaller families or minority populations, are more likely to need an unrelated donor.

Searching extended family members beyond siblings, such as cousins and aunts or uncles, is sometimes attempted but yields diminishing returns. One study that typed extended relatives for patients with common HLA haplotypes found identical donors for only 3 out of 14 patients, and those with less common haplotype combinations fared even worse, requiring more relatives to be typed with fewer matches found.⁴ Extended family searches make the most sense when the patient carries at least one very common haplotype.

Unrelated Donor Registries and the Odds They Offer

When no family match exists, patients turn to volunteer registries like the National Marrow Donor Program (known publicly as Be The Match in the U.S.) or the international network coordinated by the World Marrow Donor Association. Registries collectively hold tens of millions of volunteers who have been HLA-typed and have agreed to donate if called. Even so, the raw probability that any two random, unrelated people share the same HLA type is extremely small, typically less than 1 in 10,000.⁵ Registries overcome this by banking millions of typed volunteers, but the coverage is uneven.

A landmark study published in the New England Journal of Medicine quantified these disparities. Among patients of European descent, about 75 percent could find an optimal unrelated donor in the U.S. registry. For patients who identified as Black of South or Central American descent, that figure dropped to 16 percent.⁶ Other analyses place the broader figure for patients of European descent at roughly 60 to 80 percent for finding a 9/10 or 10/10 matched donor.⁷ The gap exists because the registry’s composition does not mirror the genetic diversity of all patients who need transplants, and because HLA diversity itself varies across populations.

Why Ethnicity Has Such a Large Effect

HLA genes are among the most variable in the human genome. This diversity exists because having a wide range of HLA types within a population is an evolutionary advantage against infectious diseases. Populations exposed to a greater variety of pathogens over thousands of years tend to have more HLA diversity.⁸ Pathogens, in effect, push the immune system to keep inventing new recognition tools, and natural selection preserves the rare variants because they help at least some individuals survive new infections.⁹

This means that populations with long histories of pathogen exposure, particularly those in sub-Saharan Africa and South Asia, tend to carry a wider array of HLA types. A study of the U.S. donor registry found high levels of HLA genetic divergence among 21 detailed racial categories, with especially broad diversity among Asian-American populations.¹⁰ India, for example, shows an extent of HLA diversity comparable to Africa, reflecting waves of migration and admixture over millennia.¹¹

Greater diversity within a population means each individual’s HLA type is more likely to be uncommon, so a larger registry is needed to cover the same proportion of patients. This is compounded in the United States by the fact that self-identified racial and ethnic categories do not map neatly onto genetic ancestry. A study of the National Marrow Donor Program found that people identifying as Black or African American had European genetic ancestry ranging broadly, and those identifying as Hispanic or Latino showed widely varying proportions of European and Amerindian ancestry even among individuals who described themselves the same way.¹² This makes it even harder to predict match likelihood from self-reported identity alone.

How Registry Size Affects Your Chances

Intuitively, a bigger registry should mean a better chance of finding a match. That is true, but only up to a point. The relationship between registry size and match probability follows a curve that rises steeply in the middle range and then levels off. At very small registry sizes, adding donors barely moves the needle because the odds of any individual being the right match are so low. At very large sizes, most of the common HLA types are already well covered, and adding more donors mostly duplicates types already present.¹³

The practical implication is that increasing a registry’s diversity matters more than just increasing its size. A registry of five million donors from one ethnic background will still fail patients from underrepresented populations. Hong Kong’s experience illustrates this neatly: with 100,000 donors, the registry could find an 8/8 match for about 45 percent of patients and a 7/8 match for about 65 percent. Doubling the registry to 200,000 pushed those numbers to 54 percent and 73 percent.¹⁴ Gains exist, but they are incremental, and the donors who matter most for closing the gap are those from underrepresented groups carrying rare HLA types.

What Happens When the Match Is Imperfect

Not every mismatch carries the same clinical weight. Mismatches at certain HLA positions cause more trouble than others. Disparities at HLA-A, -B, -C, -DRB1, -DQB1, and -DPB1 all increase the risk of severe acute graft-versus-host disease, but mismatches at HLA-C and HLA-DPB1 have a distinctive feature: while they raise GVHD risk, they are also associated with lower relapse rates, probably because the donor immune cells are more aggressive against any remaining cancer.¹⁵ This tradeoff complicates donor selection, because a slightly mismatched donor might sometimes be strategically preferable to a perfect match, depending on the disease being treated.

The level of surface expression also matters. When the mismatched HLA molecule is expressed at lower levels on the cell surface, the immune reaction tends to be milder. Specific alleles carry outsized risk. For example, a patient mismatched at HLA-C*14:02 showed the highest risk of severe acute GVHD and transplant-related mortality among all HLA-C mismatched alleles studied.¹⁶ Transplant teams now use databases of “permissive” versus “nonpermissive” mismatches to choose the safest available donor rather than simply counting the number of mismatched loci.

Alternatives When No Full Match Exists

For patients without a matched sibling or a well-matched unrelated donor, the transplant field has developed several alternatives that have dramatically expanded access over the past decade.

Haploidentical transplants use a half-matched family donor, typically a parent, child, or sibling who shares exactly one HLA haplotype. Nearly every patient has at least one haploidentical donor available, since biological parents and children are by definition half-matched. The challenge has always been the ferocious immune reaction caused by the mismatched half. A breakthrough came with the use of high-dose cyclophosphamide given after transplant, which selectively kills the activated immune cells driving graft-versus-host disease while sparing the transplanted stem cells. A systematic review and meta-analysis found that haploidentical transplants using this approach had similar overall survival compared to matched unrelated donors, and actually had better outcomes than mismatched unrelated donors.¹⁷ A multicenter trial of mismatched unrelated donor transplants using the same post-transplant cyclophosphamide approach reported one-year overall survival of 76 percent.¹⁸

These advances have tremendous potential for reducing the racial disparities that have long plagued transplant medicine.¹⁹ If a half-matched family member can produce outcomes comparable to a fully matched unrelated donor, then the desperate search through a registry that may not contain your HLA type becomes less of a bottleneck.

Cord blood transplants offer yet another option. Umbilical cord blood, collected at birth and banked, contains stem cells that are immunologically immature and therefore more tolerant of HLA mismatches. Studies have shown that some degree of HLA mismatch is acceptable with cord blood units, making them a viable option for patients who lack a matched adult donor.²⁰ Cord blood matching is evaluated at a less stringent level than adult donor matching, typically looking at six HLA markers rather than eight or ten.²¹

Beyond HLA: Other Factors That Influence Donor Selection

Even among equally well-matched donors, transplant teams weigh additional characteristics. Donor age is one of the strongest non-HLA factors. A study of the National Marrow Donor Program found that donor age was the only donor trait significantly associated with both overall and disease-free survival.²² Another study found that for patients with chronic myeloid leukemia receiving unrelated donor transplants, donor age 36 or younger and HLA matching were the two factors that maintained significance for survival in multivariate analysis.²³ The practical takeaway: when a patient has multiple equally matched donors to choose from, younger donors tend to produce better outcomes.

Cytomegalovirus (CMV) status adds another layer. CMV is a common virus that most people carry without symptoms, but it can reactivate dangerously in immunosuppressed transplant recipients. A recent large study found that when both the donor and patient were CMV-positive, older donors and HLA-mismatched donors had significantly worse outcomes. But when both were CMV-negative, those same disadvantages essentially disappeared.²⁴ This suggests that a younger, CMV-negative, slightly mismatched donor might sometimes be a better choice than an older, CMV-positive, perfectly matched one.

What Donation Looks Like for the Donor

If you are on a registry and get called as a match, understanding what the process involves can matter for your decision to follow through. There are two collection methods: bone marrow harvest and peripheral blood stem cell (PBSC) collection. For bone marrow harvest, donors undergo general anesthesia and have marrow extracted from the back of the pelvic bone through a needle. For PBSC collection, donors receive injections of a growth factor for several days that pushes stem cells from the marrow into the bloodstream, after which the cells are collected through a process similar to blood donation.

Both methods cause temporary symptoms. In the National Marrow Donor Program’s experience, pain was the most reported symptom for both groups. About 82 percent of marrow donors reported back or hip pain at the collection site, while 97 percent of PBSC donors reported bone pain at various sites during the growth factor injections. Fatigue was the second most common complaint for both. PBSC donors typically recovered in about a week, while marrow donors took a median of about two to three weeks.²⁵ A clinical trial comparing the two approaches confirmed these differences, with marrow donors more likely to report skeletal pain and fatigue at one week while PBSC donors had greater temporary changes in blood counts.²⁶ Serious adverse events were uncommon in both groups, though slightly more frequent among marrow donors.

Why HLA Diversity Exists in the First Place

It might seem like a design flaw that our immune markers are so varied, making transplant matching so difficult. But that variety is exactly what keeps human populations resilient against infectious disease. Each HLA type is specialized to recognize a different set of pathogen fragments. A population with many HLA types is harder for any single pathogen to overwhelm, because even if one HLA variant fails to detect an emerging infection, other variants will catch it. This is why HLA genes are under what geneticists call balancing selection: rather than one “best” version winning out, natural selection actively preserves many versions at once.

The geographic pattern confirms this. Populations in tropical regions with high pathogen diversity tend to have more HLA variation, while populations in temperate regions with fewer pathogens tend to have somewhat less. The tradeoff is clear when it comes to medicine: the very diversity that protects us from pandemics also makes it harder to find transplant matches, and it makes the search especially hard for populations with the richest immunological heritage.