What Blood Type Will This Child Have?
Parent 1
Advanced: I know this parent's genotype
Leave both on “use population estimate” unless a genetic test has told you the alleles. Most calculators force you to pick one; here it is genuinely optional.
Parent 2
Advanced: I know this parent's genotype
Leave both on “use population estimate” unless a genetic test has told you the alleles.
A family tree or an existing child is stronger evidence than any allele frequency. Fill in either block and the estimate tightens. It is applied as Bayes evidence, not as an override.
Evidence A — this parent's own parents (grandparents of the child)
Parent 1's parents
Parent 2's parents
Both grandparents of one parent are needed before the block activates, because one grandparent alone cannot narrow a genotype.
Evidence B — a child already born, or a test result
This runs backwards. An O child proves both parents carry an O allele. An Rh negative child from two Rh positive parents proves both are Dd, which turns the next child into a clean 1 in 4.
| Blood type | Likelihood | Share | Odds | Status |
|---|---|---|---|---|
| Total | 100.00% | |||
ABO and RhD sit on different chromosomes (9 and 1) and are inherited independently, so each gets its own 4×4 square. The odds table above multiplies them.
Whole red cells only. Plasma compatibility runs the other way round, and a hospital cross match is the only authority on whether a specific unit is safe for a specific person. Reference data only — not a transfusion decision tool.
What this calculator actually computes
Most blood type calculators ask you for a genotype. A lab report never gives you one. It gives you a label, A positive or O negative, and that label is a phenotype: a description of what is sitting on the surface of your red blood cells. Underneath, almost every person carries two copies of the ABO gene and two copies of the RhD gene, and the label only tells you what those two copies do together.
So there are two numbers where a calculator usually wants one. A person typed as A is either AA or AO. A person typed as positive is either DD or Dd. Which one it is, a blood test cannot tell you, and neither can a family tree. It changes the answer. Two parents both labelled A positive are labelled positive, so at least one D allele is present in each of them, but a Dd parent carries a hidden d allele. The odds of an Rh negative child are not zero, and they are not one in four either.
This calculator takes the labels you actually have and propagates that uncertainty instead of hiding it. Each parent enters as a probability distribution over the genotypes that could produce their label. The child's distribution is the sum over every possible parental pairing, weighted by how likely that pairing is. Nothing gets rounded into a guess.
You can add evidence, and the maths tightens with it. If you know your parents' blood types, that is another generation of inheritance to work backwards from. If a child has already been born, or a test has already been done, that observation is stronger than any population average and it is applied as evidence rather than as a footnote. Two parents labelled A and B who already have an O child are no longer a maybe: an O child has to receive an O allele from each side, so both parents are carriers, and the answer changes from a range to a fact.
How the numbers get there
ABO has three alleles. A and B are codominant, so a person carrying both is typed AB. O is recessive, so it only shows when both copies are O. That gives six genotypes behind four labels.
The probabilities start from population allele frequencies. Those are the one assumption in this tool that is not measured, so they are stated here in full and you can check them: A at 0.2605, B at 0.0704, O at 0.6691. Squaring and pairing those under Hardy-Weinberg equilibrium gives genotype frequencies, and adding them up gives the phenotype split that makes the choice defensible: O 44.8 percent, A 41.6 percent, B 9.9 percent, AB 3.7 percent. That is the blood group distribution usually published for populations of European ancestry, which is where the rest of the site's health and genetics figures are drawn from too.
Those genotype frequencies are then conditioned on the label you actually entered, and conditioning is what matters. Under these numbers about 41.6 percent of people are typed A, and of those, 83.7 percent are AO and only 16.3 percent are AA. If your family comes from somewhere else, that prior is wrong for you, which is exactly why the grandparent boxes exist: enter real family types and the population average stops being used for that parent.
RhD is simpler. D is dominant over d, so a person typed positive is either DD or Dd and a person typed negative is always dd. This tool assumes a d allele frequency of 0.118, which puts Rh negative at about 1.4 percent, again the figure usually published for populations of European ancestry. Under it, a person typed positive is Dd about 21.1 percent of the time and DD the rest. Rh negative is much more common in some other populations, so for those this prior understates the chance of a Dd carrier, and a family tree will correct it.
The two genes sit on different chromosomes, chromosome 9 for ABO and chromosome 1 for RHD, so they are inherited independently and their probabilities multiply. That is why two A positive parents with no other information come out at 81.6 percent A positive, 17.3 percent O positive, 0.9 percent A negative and 0.2 percent O negative, instead of the 100 percent A positive that a naive answer gives.
One detail causes more wrong answers than any other. A DD parent masks the d allele completely, so a DD crossed with a Dd parent cannot produce an Rh negative child. Only Dd crossed with Dd can. If a calculator reports one in four for any two Rh positive parents, it has not applied dominance correctly.
What this model does not know
No calculator, including this one, predicts a baby's blood type. It gives you odds, and the odds are only as good as the assumptions underneath.
The Bombay phenotype, hh, breaks the whole framework. Those people look like group O on a routine test because they do not build the H antigen that A and B are built on top of, yet they carry real A or B alleles. Standard ABO inheritance does not describe them. cis-AB, a genetic oddity where A and B are fused into one allele, breaks the codominance assumption in a different way. Weak D and partial D are clinically meaningful variants of the RhD antigen that a label reading simply positive does not distinguish.
The population frequencies are averages. Actual frequencies differ by ancestry, and a well-kept family tree beats any average. Enter your parents' types and the calculator stops relying on the average for that parent. Two parents of the same phenotype can produce different looking results simply because their families are different.
This is a genetics teaching tool. It is not a diagnostic device, it does not test anybody's blood, and it cannot tell you whether a transfusion is safe. For that, a laboratory and a clinician are the only answer.
Why not ask a chatbot
Chatbots are confidently wrong about blood type inheritance in ways that are hard to catch. The common failure is treating a phenotype as a genotype, which is how two type A positive parents end up with a 100 percent A positive answer. The second is getting dominance backwards, which is how an Rh positive parent ends up contributing a recessive allele they do not have. The third is mixing up codominance with blending, which quietly turns A into AA across a generation. All three produce fluent sentences, and all three are wrong in the same direction: too confident.
The arithmetic here is not a prompt. It is a probability model over 36 ABO genotype pairings and 9 RhD pairings, with the arithmetic and the closed-form results checked against a test gate that has to go green before this page ships. Two parents both labelled O return a certain O with no hedging. Two parents both labelled O negative return a certain O negative. Two A positive parents return a distribution, because the honest answer is a distribution. A chatbot asked the same question three times will give you three different certainties.
Sources and references
- Dean L. Blood Groups and Red Cell Antigens. Blood Bank Transfusion Services, NCBI Bookshelf NBK2261 (the ABO chapter is NBK2267). The standard reference for ABO and Rh(D) antigens, genotypes and inheritance. ncbi.nlm.nih.gov/books/NBK2261
- MedlinePlus Medical Encyclopedia, Blood typing. US National Library of Medicine. Patient-facing explanation of the ABO and Rh systems and of what a blood type label does and does not tell you. medlineplus.gov/ency/article/003345.htm
- MedlinePlus, Erythroblastosis Fetalis / Rh Incompatibility, and the Medical Encyclopedia article on Rh incompatibility. The clinical background to the anti-D note above. medlineplus.gov/rhincompatibility.html · ency/article/001600.htm
- AABB Technical Manual, chapters on ABO and Rh(D) discrepancies and on transfusion compatibility. The source for the red blood cell compatibility table.
- Hh blood group (Bombay phenotype). Background on the hh phenotype named in the limitations section above.
The allele frequencies used as the starting prior are listed in full in “How the numbers get there” above, and every other number on this page is arithmetic on top of them.
Frequently Asked Questions
What blood type will my baby have?
That depends on both parents' genotypes, which a blood type label does not reveal. Type A means AA or AO, and Rh positive means DD or Dd. Enter both parents' labels above and the calculator weights every possible parental pairing by how likely it is, then reports the odds for each of the eight phenotypes. Certain answers show up as certain: two type O parents always produce a type O child, and two Rh negative parents always produce an Rh negative child.
Can two type A positive parents have a type A negative child?
Yes, but not often. An Rh positive parent is DD roughly 78.9 percent of the time and Dd roughly 21.1 percent of the time, using a recessive d allele frequency of 0.118. Only a Dd crossed with a Dd can produce an Rh negative child, because a DD parent masks the recessive allele completely. Under the standard allele model that gives two type A positive parents an A negative child about 0.92 percent of the time and an O negative child about 0.20 percent of the time. A calculator that reports one in four for any two Rh positive parents has not applied dominance correctly.
What blood type can two type A parents have?
Only A or O. The ABO gene has three alleles: A and B are codominant, O is recessive. Two type A parents carry no B allele at all, so a B or AB child is impossible. The split between A and O depends on how many of the parents are heterozygous. Without any other information the odds are about 82.5 percent A and 17.5 percent O, because a type A person is AO about 83.7 percent of the time. If they already have an O child, both parents are proven heterozygous AO and the answer becomes 75 percent A and 25 percent O with no uncertainty.
How does the Punnett square work for blood type?
Each parent contributes one allele, so you build a grid with one parent's two alleles across the top and the other's down the side. For ABO, A and B are codominant so AB shows both antigens, while O is recessive and only appears as OO. For RhD, D is dominant over d, so dd is the only Rh negative genotype. Because the ABO gene sits on chromosome 9 and the RHD gene on chromosome 1, the two are inherited independently, so this calculator shows a separate 4×4 square for each and multiplies their probabilities rather than pretending the pair fits in one grid.
What is the difference between a blood type and a blood genotype?
The blood type is the phenotype, what a lab report prints: A, B, AB or O, positive or negative. The genotype is the pair of alleles underneath: AA, AO, BB, BO, AB or OO, and DD, Dd or dd. Most calculators ask for the genotype, which is information a routine blood test does not give you, so they either force you to guess or silently assume homozygous parents. This one asks only for the labels you have and carries the uncertainty through the arithmetic.
Can a blood type calculator tell me if my baby will have a certain blood type?
It can give you odds, not a prediction. Blood type inheritance follows ordinary Mendelian ratios once genotypes are known, but genotypes are not observable from a routine test. An A positive child is not proof of anything on its own unless the parents' phenotypes already restrict the possibilities: an O child rules out every parental pairing where either parent has no O allele, and an Rh negative child from two Rh positive parents proves both are Dd. Where an observation changes the odds, this calculator shows the change.
Is there such a thing as the Bombay blood group, and does this calculator handle it?
Yes, and no. The Bombay phenotype, hh, occurs when a person carries two recessive copies of the H gene and never builds the H antigen that A and B sit on, so routine testing types them as O even though they carry real A or B alleles. Standard ABO inheritance does not describe them, so this calculator does not model it. It also does not model cis-AB, weak D, partial D or A1/A2 subgroups. These are rare, they need specialist testing, and pretending to handle them would be worse than saying so.
Can AI chatbots calculate blood type inheritance correctly?
Chatbots are wrong in three specific, consistent ways. First, they treat a phenotype as a genotype, which is how two type A positive parents get a confident 100 percent type A positive answer. Second, they get dominance backwards, so an Rh positive parent contributes a recessive allele they do not carry and the odds come out as one in four. Third, they blend instead of codominate, so A quietly becomes AA across a generation. Each error makes the answer look more certain than it is, and a chatbot asked three times gives three different certainties. This calculator runs a fixed probability model with the arithmetic checked against a test gate before the page ships.
Why does an Rh negative mother and an Rh positive father matter?
An Rh negative mother can carry Rh positive red blood cells during pregnancy and respond by producing antibodies against the D antigen. This is Rh isoimmunisation, and it can affect a later pregnancy. Standard practice is to offer anti-D immunoglobulin to Rh negative mothers, which is why the pairing is flagged here rather than left as a number. This is a general note, not medical advice, and the timing and dosing are decisions for a clinician.
Which blood types can donate to which?
For red blood cells with ABO and Rh only, O negative is the universal donor and AB positive the universal recipient, but cross matching in a hospital is always the authority. Within ABO alone, O can donate to everyone and AB can receive from everyone. Rh follows a separate rule: Rh positive recipients can take Rh positive or Rh negative, and Rh negative recipients can only take Rh negative. Plasma reverses the pattern, which is why group compatibility tables are labelled by product. Nothing on this page replaces a laboratory cross match.
Is my blood type information private?
Yes. Everything runs as JavaScript in your browser. No blood type, no family detail and nothing about your children is sent to any server, because there is no server involved. The only stored data is your current selection in your own browser local storage so the page is still there when you come back, plus the file you download if you export one. There is no account, no tracking and no analytics tied to what you type.
Comments & Ratings