23andMe does not include a blood type report in its standard results. You will not find a page in your dashboard telling you whether you are type A, B, AB, or O. However, the raw genetic data that 23andMe collects does contain the key DNA variants needed to figure out your ABO blood type, and with a little effort or the right third-party tool, you can extract that information yourself. The gap between what 23andMe sequences and what it chooses to display is the real story here.
What 23andMe Actually Tests
23andMe uses a genotyping chip that reads hundreds of thousands of specific positions in your DNA, called single nucleotide polymorphisms, or SNPs. Several of those SNPs sit within the ABO gene on chromosome 9, which is the gene that determines your blood type. Researchers have identified a small set of these variants that reliably predict ABO type. Four SNPs are commonly used for this purpose: rs8176719 (an exon-6 deletion that tags the O1 type), rs41302905 (which tags O2), rs8176746 (which distinguishes B), and rs56392308 (which identifies the A2 subtype).1PLoS ONE. A SNP panel and online tool for checking genotype concordance through comparing QR codes So the data is there in your raw file. 23andMe just hasn’t built an interface to interpret it for you.
If you download your raw data file from 23andMe’s settings page, you’ll get a text file containing your genotype at every tested position. You can search that file for the relevant SNP identifiers and, with some knowledge of which allele combinations correspond to which blood types, work out your ABO group manually. Most people, understandably, would rather not do that by hand.
Third-Party Tools That Do the Work for You
Several tools and web services have been designed to take raw genetic data from consumer testing companies and predict blood type. One academic tool called BOOGIE was specifically built to predict blood groups from sequencing data. When tested on a 23andMe dataset, it achieved about 91% accuracy, correctly calling 46 out of 51 type O individuals, all 44 type A individuals, 18 out of 21 type B individuals, and 4 out of 5 type AB individuals.2PLoS ONE. BOOGIE: Predicting Blood Groups from High Throughput Sequencing Data Most of the errors came from difficulty detecting one particular deletion variant, not from a fundamental flaw in the approach. The researchers noted that using just a handful of localized SNPs, rather than full genome data, can still be effective for ABO prediction.
Another tool, PAGEANT, was developed to let people explore traits from their personal genomes, including blood type. It treats ABO blood type as a straightforward categorical trait that can be determined from the presence of specific target variants in raw data.3Nucleic Acids Research. PAGEANT: personal access to genome and analysis of natural traits Community-built tools on platforms like Promethease and SNPedia can also parse 23andMe data for blood type. These vary in how user-friendly they are, but they all rely on the same underlying principle: reading a handful of SNPs in the ABO gene from your downloaded raw data file.
Why the Accuracy Isn’t Always Perfect
You might wonder why genetic prediction of something as basic as blood type is only around 91% accurate from consumer genotyping data, when hospitals determine it with near-total certainty from a drop of blood. The answer comes down to how different technologies work. A serological blood test physically mixes your blood with antibodies and watches what clumps together. It’s testing the final product directly. Genetic prediction, by contrast, is reading the recipe and inferring the outcome.
Consumer genotyping chips only test pre-selected spots in the genome. If the chip doesn’t include one of the critical ABO variants, or if the variant is an insertion/deletion that the chip struggles to read cleanly, the prediction can go wrong. The BOOGIE researchers found that most of their errors on 23andMe data traced back to one specific deletion variant that was either missing or misread in a few samples.4PLoS ONE. BOOGIE: Predicting Blood Groups from High Throughput Sequencing Data When researchers use whole-genome sequencing, which reads essentially everything rather than just selected spots, accuracy jumps dramatically. One algorithm developed for clinical-grade sequencing data achieved over 99% concordance with serological typing across hundreds of genomes.5The Lancet. Automated typing of red blood cell and platelet antigens: a whole-genome sequencing study Another tool called RBCeq reached about 99% accuracy across 21 blood group systems in over 400 samples.6PubMed Central. RBCeq: A robust and scalable algorithm for accurate genetic blood typing
The takeaway: a blood type derived from 23andMe raw data is a reasonable estimate, but it’s not something you should hand to a hospital before a transfusion. If you need your blood type for medical purposes, a standard serological test from a lab or blood donation center is the only reliable approach.
Why 23andMe Doesn’t Just Tell You
This is a question that frustrates a lot of 23andMe customers. The company clearly has the relevant data, so why not display it? 23andMe hasn’t given an official public explanation, but a few factors likely play into the decision. First, the company curates its health and trait reports carefully, typically only releasing reports that meet a certain threshold of clinical validity and regulatory review. Since consumer genotyping has a small but real error rate for blood type, putting a definitive-looking “Your blood type is A+” on someone’s dashboard could create problems if that person later relied on it medically.
Second, the ABO gene alone doesn’t tell the full story. The Rh factor (the “positive” or “negative” part of blood type) is determined by a separate gene, RHD, and accurately calling Rh status from genotyping chip data presents its own challenges. Displaying only ABO type without the Rh component would feel incomplete, and displaying both requires confidence in both calls. Third, blood type simply may not be a priority for 23andMe’s product team compared to ancestry features and disease risk reports, which drive their business.
Beyond ABO and Rh
When most people say “blood type,” they mean the ABO group and whether they’re Rh positive or negative. But human blood cells actually carry dozens of antigen systems. Kell, Duffy, Kidd, and MNS are among the more clinically relevant ones, especially for people who need repeated blood transfusions or who are pregnant and at risk for hemolytic disease of the newborn. Genetic sequencing can now type many of these systems simultaneously, which is a significant advance over older methods that required panels of antibody reagents for each system.
The high-accuracy algorithms mentioned earlier, like the one published in The Lancet and RBCeq, were designed to type red blood cell antigens across multiple systems, not just ABO.7The Lancet. Automated typing of red blood cell and platelet antigens: a whole-genome sequencing study 8PubMed Central. RBCeq: A robust and scalable algorithm for accurate genetic blood typing This kind of comprehensive genetic blood typing is moving toward clinical use, particularly for patients with rare blood types or those in populations where finding compatible donors is difficult. 23andMe’s chip data, however, is unlikely to cover enough of these minor blood group variants to be useful for extended typing.
Secretor Status and What It Means
One related trait that some third-party tools can pull from raw data is secretor status. This refers to whether you express your blood type antigens not just on your red blood cells but also in body fluids like saliva, tears, and the mucus lining your gut. About 80% of people are “secretors,” meaning they do express these antigens in fluids, while roughly 20% are “non-secretors” who don’t.9PubMed Central. Expression of the gene encoding secretor type galactoside 2 alpha fucosyltransferase (FUT2) and ABH antigens in patients with oral lesions
Secretor status is controlled by the FUT2 gene, and certain mutations in that gene shut off the enzyme that moves blood group antigens onto mucosal surfaces. This distinction matters more than you might expect. The FUT2 gene has been studied as a susceptibility marker for various infectious diseases, because the presence or absence of blood group antigens in your gut lining affects which pathogens can latch on.10PubMed Central. FUT2 gene as a genetic susceptible marker of infectious diseases: A Review Non-secretors, for instance, appear to have some protection against norovirus, which exploits these surface antigens as a way to infect gut cells. 23andMe genotypes SNPs within the FUT2 gene, so some users can also find their secretor status through the same raw data download process.
Blood Type and Disease Risk
Part of the reason people want to know their blood type from 23andMe is curiosity about health implications. There is genuine science behind some of these associations, though the effect sizes tend to be modest. The strongest evidence links non-O blood types (A, B, and AB) with a slightly higher risk of cardiovascular disease. A large analysis combining data from the Nurses’ Health Study and the Health Professionals Follow-up Study found that, compared to people with type O blood, those with type AB had about a 23% higher risk of coronary heart disease, those with type B had about a 15% higher risk, and those with type A had a roughly 6% higher risk. Overall, about 6% of coronary heart disease cases in the study population were statistically attributable to having a non-O blood group.11PubMed Central. ABO blood group and risk of coronary heart disease in two prospective cohort studies Genome-wide association studies have also flagged the ABO locus as connected to thrombosis and various cardiovascular risk biomarkers.12PubMed Central. ABO Blood Groups and Cardiovascular Diseases
These are real associations, but they’re small enough that your blood type shouldn’t change your behavior. The standard advice for heart health, regular exercise, good diet, not smoking, applies equally whether you’re type A or type O. Blood type is one tiny variable in a landscape dominated by lifestyle factors, cholesterol levels, blood pressure, and family history of heart disease.
The Blood Type Diet Myth
If you’ve gone looking for your blood type, there’s a decent chance someone has told you about the “blood type diet,” which claims that people with different blood types should eat fundamentally different foods. Type O should eat high-protein, type A should be vegetarian, and so on. The idea was popularized in a bestselling book in the late 1990s and has remarkable staying power online.
The science, however, does not support it. A systematic review published in the American Journal of Clinical Nutrition found no evidence to validate the claimed health benefits of blood type diets.13PubMed. Blood type diets lack supporting evidence: a systematic review A separate study that actually tested the hypothesis by looking at whether adherence to specific blood-type diets improved cardiometabolic markers found something telling: some of the recommended diets did produce health benefits, but the benefits had nothing to do with the person’s blood type. People who followed the “type A” diet (which is largely plant-based) saw improvements regardless of whether they were actually type A.14PLOS ONE. ABO Genotype, ‘Blood-Type’ Diet and Cardiometabolic Risk Factors In other words, eating more vegetables helps everyone. That’s not a blood-type insight; it’s basic nutrition.
Privacy Considerations With Blood Type in Raw Data
There’s a dimension to all of this that doesn’t always occur to people when they download their raw genetic data or upload it to third-party tools. Blood type is a phenotypic trait that can be predicted from just a handful of SNPs, and researchers have demonstrated that publicly available SNP-phenotype associations can be used to cross-link individuals between genetic and phenotypic datasets. In one study, researchers predicted traits including blood type from as few as 20 SNPs and used that information to re-identify individuals across separate databases.15PubMed Central. Assessing Privacy Vulnerabilities in Genetic Data Sets: Scoping Review
This doesn’t mean you should panic about uploading your raw data to a blood type calculator. But it’s worth understanding that genetic data, once shared, can reveal more than you intended. Blood type on its own is not very identifying since millions of people share each type. Combined with other predicted traits like eye color, hair color, and skin color, though, the set of characteristics starts to narrow. If you’re thinking about uploading your raw 23andMe data to a third-party service, checking the service’s privacy policy is a reasonable precaution.
Why ABO Polymorphism Exists at All
One of the more interesting puzzles in human genetics is why we have blood type variation in the first place. This isn’t just a human quirk. The same ABO variants have been found in other primate species, and the genetic polymorphism appears to have been maintained for at least 20 million years.16PubMed Central. Ancestry runs deeper than blood: the evolutionary history of ABO points to cryptic variation of functional importance When a genetic variant persists across that many species and that much evolutionary time, it usually means natural selection is actively preserving it, a process called balancing selection.
The leading theory is that gut pathogens are the driving force. Different blood type antigens appear on the surface of gut cells, and different pathogens have evolved to exploit specific antigens for entry. A population with a mix of blood types is harder for any single pathogen to sweep through, so the variation itself is what’s protective. The global distribution of blood types loosely tracks historical disease pressures, with type O being most common in regions with heavy malaria exposure and type B being more frequent in parts of Central and East Asia.17PubMed Central. A brief history of human blood groups This is also probably why blood type keeps showing up in disease association studies: the antigens aren’t just passive markers, they’re active participants in how pathogens interact with your cells.
The Simplest Way to Learn Your Blood Type
If you’ve read this far and you’re still hoping 23andMe will just tell you, the fastest and most reliable route remains decidedly low-tech. Donate blood. Nearly every blood donation center will tell you your ABO type and Rh factor after your first donation, and the result comes from a direct serological test with essentially no margin of error. Many primary care doctors can also order a blood type test, though this often isn’t covered as a routine screening and may cost a small fee depending on your insurance. Home blood typing kits exist too, typically using a card with dried antibodies that you apply a blood drop to, and they’re reasonably accurate when used correctly.
For people who specifically want to dig into their 23andMe data, the raw file approach works and the available tools are getting better over time. Just keep in mind that about one in ten predictions from consumer genotyping data may be off, particularly for less common blood types like B and AB where the sample sizes in validation studies are smaller. If the result matters for anything medical, confirm it with a proper lab test.