Audited —·Last updated 29 Jul 2026·6 citations·Tier 2·0 uses

Blood Type Inheritance Calculator

Free blood type inheritance calculator. Enter both parents' ABO and RhD genotypes for exact child blood group probabilities. Not a paternity test.

Blood Type Inheritance Calculator

Parent 1 — ABO genotype
Parent 2 — ABO genotype
Parent 1 — RhD genotype
Parent 2 — RhD genotype
Blood group A
25.00
Percentage of offspring in ABO group A (genotype AA or AO). Exact under the classical three-allele model with the genotypes you entered — this is deterministic Mendelian arithmetic, not an estimate, and not a statement about any particular child.
Blood group B
25.00
Blood group AB
25.00
Blood group O
25.00
RhD positive
75.00
RhD negative
25.00
ABO ratio
1 O : 1 A : 1 B : 1 AB
Possible blood types
O+, O−, A+, A−, B+, B−, AB+, AB−
ABO groups excluded
none — all four ABO groups are possible
Most likely blood type
O positive or A positive or B positive or AB positive
Probability of that type
18.75

Background.

This calculator works out which blood groups two parents can have children with, and with what probability, for the ABO system and the RhD antigen together. It takes each parent's ABO genotype and RhD genotype and returns the exact percentage for groups O, A, B and AB, the percentage RhD positive and negative, the full list of possible types, and — usually the most useful output — which groups are excluded.

It asks for genotypes rather than blood types on purpose, and that choice is worth explaining because it is the difference between an honest answer and a fabricated one. A blood test tells you a phenotype. It cannot distinguish genotype AA from AO, or BB from BO, and routine anti-D typing cannot distinguish DD from Dd. A calculator that accepts 'group A' and returns probabilities has to invent a split between AA and AO, which depends on allele frequencies that vary substantially by ancestry. That would be a guess presented as arithmetic. With genotypes supplied, the result is deterministic Mendelian stoichiometry and nothing is invented. If you do not know a parent's genotype, the right move is to run each compatible option and take the union of the possible types — that set is determinate even when the probabilities are not.

The genetics is a good teaching case because ABO breaks the simple dominant/recessive pattern in two ways at once. One locus carries three alleles rather than two, giving six genotypes and four phenotypes. And A and B are codominant with each other while both are dominant over O: the NCBI summary puts it as 'the A and B alleles are codominant, and the O allele is recessive'. The A and B alleles each encode a working glycosyltransferase that adds a different sugar to the H-antigen precursor; the O allele encodes an enzyme with no function, so a person with two O alleles displays the unmodified precursor. Someone with genotype AB makes both enzymes and shows both sugars, which is what codominance means physically.

RhD is a separate story on a separate chromosome. The RHD gene at 1p34–36 encodes the D antigen; having at least one working copy makes you RhD positive. The 'd' written here is not really an allele — it is shorthand for the absence of a working RHD gene, usually an outright deletion in European populations. Because RHD and ABO are on different chromosomes they assort independently, so the calculator multiplies the two results together.

The limits belong at the front, not at the bottom. This is not a paternity test. Blood groups can occasionally exclude a claimed parent, they can never confirm one, and the exclusions themselves have documented exceptions. The cis-AB allele produces both A and B antigens from a single allele, so a cis-AB parent typed as AB can have a group-O child — a peer-reviewed case series in Annals of Laboratory Medicine describes exactly that, and puts the allele's frequency at 0.0354 % in Korea against 0.0012 % in Japan and 0.00066 % in China. The Bombay phenotype suppresses A and B expression regardless of ABO genotype. Weak D and partial D variants type positive with some reagents and negative with others. Blood chimerism, ABO subgroups such as A2 and A3, and any recent transfusion or transplant can all move a typing result. Legal or personal parentage questions need accredited DNA testing, not a Punnett square. This page is also not a transfusion tool and gives no transfusion or anti-D guidance.

What is blood type inheritance calculator?

The ABO blood group is controlled by one gene on chromosome 9 with three classical alleles: A, B and O (also written Iᴬ, Iᴮ and i). Three alleles give six genotypes, and those six genotypes produce four observable blood groups.

Genotype AA or AO gives blood group A. Genotype BB or BO gives group B. Genotype AB gives group AB. Genotype OO gives group O. The A and B alleles each encode a glycosyltransferase that attaches a different terminal sugar to the H antigen already present on the red cell membrane; the O allele carries a single-base deletion and encodes an enzyme with no activity, so group-O red cells display the unmodified H antigen. That is why A and B are codominant with each other — both enzymes work, so both sugars appear — while both are dominant over O, which contributes nothing to detect.

The RhD antigen is controlled separately, by the RHD gene at 1p34–36, alongside the closely linked RHCE gene. Carrying at least one working RHD copy makes a person RhD positive; the RhD-negative phenotype in European populations usually results from deletion of RHD altogether, and in African populations more often from an RHD pseudogene or a hybrid RHD–RHCE gene, which is why the same phenotype has different molecular causes in different ancestries. Because RHD is on chromosome 1 and ABO on chromosome 9, the two are inherited independently and the probabilities multiply.

A full blood type combines the two — 'A positive' means ABO group A plus at least one working RHD. The Rh system actually contains more than fifty antigens, of which D, C, c, E and e are the commonly typed ones; D is singled out because it is by far the most immunogenic. This page models D only, and models it as a simple two-state dominant character, which is the standard teaching model and not the full molecular picture.

How to use this calculator.

  1. Work out each parent's ABO genotype, not just their blood type. Group AB and group O are unambiguous (AB and OO). Group A is AA or AO and group B is BB or BO, and a blood test cannot tell you which.
  2. If a group-A or group-B genotype is unknown, run the calculator twice — once for the homozygous option and once for the heterozygous one — and take the union of the possible types. The set of possible outcomes is determinate even when the probabilities are not.
  3. Do the same for RhD. An RhD-negative person is dd for certain; an RhD-positive person may be DD or Dd, and routine typing cannot distinguish them.
  4. Read the four ABO percentages and the two RhD percentages. Each set adds to exactly 100%.
  5. Look at 'ABO groups excluded' before anything else. Exclusion is the only direction in which blood groups carry real information, and it is what makes the classic textbook results interesting — for example, two group-O parents can only have group-O children.
  6. Use 'possible blood types' for the combined ABO + Rh answer, and the most-likely output for the single highest-probability combination. Where several combinations tie, all of them are listed rather than one being picked at random.
  7. Do not use any of it to settle a parentage question. Blood groups can sometimes exclude, never confirm, and the exclusions have documented exceptions — accredited DNA testing is the only correct route.

The formula.

P(group) = cells⁄4 ; P(type) = P(ABO) × P(RhD)
A, B · codominant alleles, both dominant over OO · recessive, encodes a non-functional enzymeD · working RHD gene, dominantd · absence of a working RHD gene, not a real allele4 · cells in each 2 × 2 grid

Two independent Punnett squares, multiplied.

ABO first. Each parent passes on one of its two ABO alleles, each with probability ½, so the 2 × 2 grid holds four equally likely cells and P(genotype) = cells ⁄ 4. The genotype is then mapped to a blood group by the dominance rules: carrying both an A and a B allele gives group AB (codominance — both enzymes are made); carrying an A but no B gives group A; carrying a B but no A gives group B; and only OO gives group O.

RhD second, on the same 2 × 2 pattern: any offspring carrying at least one D types RhD positive, and only dd types negative.

The two are then combined. ABO sits on chromosome 9 and RHD on chromosome 1p34–36, so they assort independently and the joint probability is simply the product:

P(A positive) = P(group A) × P(RhD positive)

For the worked example that is 25 % × 75 % = 18.75 %.

ROUNDING STAGE. Rounding happens only at the final return, to ten decimal places. The ABO and RhD probabilities are exact counts out of four, so the only values reachable are 0, 25, 50, 75 and 100; the joint probabilities are products of those and are multiples of 6.25. Nothing is rounded part-way through and there is no threshold anywhere in the calculation.

WHY GENOTYPES AND NOT BLOOD TYPES. Because the map from genotype to phenotype loses information and cannot be inverted. Group A means AA or AO; group B means BB or BO; RhD positive means DD or Dd. To turn a phenotype back into a probability distribution over genotypes you would need population allele frequencies, which differ by ancestry and would be an assumption smuggled into an answer that looks exact. This page refuses that trade and asks for the genotype instead, then tells you what to do when you do not have one.

WHERE THE MODEL BREAKS. The cis-AB allele carries a single enzyme with both A and B activity, so both antigens are inherited from one parent on one allele; a cis-AB parent typed as AB can therefore have a group-O child, which the classical model says is impossible. The Bombay phenotype (homozygous null at the H locus) leaves no H precursor for the A or B enzymes to modify, so such a person types as group O regardless of their ABO genotype and can pass on A or B alleles. ABO subgroups such as A2, A3 and Ax type weakly and are sometimes misread as O. Weak D and partial D type inconsistently between reagents. Blood chimerism, recent transfusion and haematopoietic transplant all change what a typing test reports. None of these are modelled here; all are named because each one can turn a confident-looking exclusion into a wrong one.

INVALID DOMAIN. There is no singularity — denominators are the constants 4 and 100. The calculator rejects any ABO genotype outside {AA, AO, BB, BO, AB, OO} and any RhD genotype outside {DD, Dd, dd}, including a bare blood type such as 'A' entered where a genotype is required, with a message under the offending field. A blood group with zero probability is reported as 0 % and listed under the excluded groups, which is a correct answer rather than an error.

A worked example.

Example

The classic demonstration: a group-A parent and a group-B parent who between them can produce children of all four blood groups, including two groups neither of them has. Parent 1 has genotype AO, so passes on either an A allele or an O allele. Parent 2 has genotype BO, so passes on either B or O. The four equally likely combinations are AB, AO, OB and OO — one cell each. Mapping them to blood groups: AB is group AB, AO is group A, BO is group B, and OO is group O. So each of the four groups comes out at exactly 25 %, and the ABO ratio output reads 1 O : 1 A : 1 B : 1 AB. The 'groups excluded' output reads that none are excluded, which is the whole point of the example: nothing about these two parents rules out any ABO group in their children. RhD is a separate grid. Both parents are Dd, so both are RhD positive while each carries one non-working RHD. Their four combinations are DD, Dd, Dd and dd, so three quarters of children carry at least one working copy and type RhD positive, and one quarter are dd and type RhD negative. The calculator returns 75 % positive and 25 % negative. Two RhD-positive parents can therefore have an RhD-negative child, which surprises people and is entirely ordinary. Combining the two is a multiplication, because ABO and RHD are on different chromosomes. Each ABO group at 25 % times RhD positive at 75 % gives 18.75 %, and times RhD negative at 25 % gives 6.25 %. All eight combinations are possible — O+, O−, A+, A−, B+, B−, AB+, AB− — and the four positive ones tie at 18.75 %. The most-likely output lists all four rather than picking one, because they are genuinely equal: 'O positive or A positive or B positive or AB positive', at 18.75 %. The eight joint probabilities add to 4 × 18.75 + 4 × 6.25 = 100 %. Now the opposite case, which shows why exclusions matter more than probabilities. Set parent 1 to AB and parent 2 to OO, both dd. Parent 1 passes A or B; parent 2 can only pass O. Every child is AO or BO, so group A comes out at 50 % and group B at 50 %, while groups AB and O are both 0 % and appear under 'ABO groups excluded'. Both parents are RhD negative, so all children are RhD negative and only A− and B− are possible. Notice what happened: an AB parent and an O parent can have no child whose blood group matches either of them. That exclusion — an AB parent cannot have a group-O child — is exactly the rule that the cis-AB allele breaks. Chun and colleagues, writing in Annals of Laboratory Medicine in 2019, document precisely 'the birth of an O child from an AB mother' where the mother carries cis-AB rather than ordinary AB. The allele is rare, at 0.0354 % in Korea and far rarer elsewhere, but it is real, and it is why a blood-group exclusion is never sufficient evidence about parentage.

parent2 Abo GenotypeBO
parent2 Rh GenotypeDd
parent1 Rh GenotypeDd
parent1 Abo GenotypeAO

Frequently asked questions.

Can two group-O parents have a group-A child?
Under the classical model, no. Group O means genotype OO, so each parent can pass on only an O allele and every child must be OO — the calculator returns 100 % group O and lists A, B and AB as excluded. That said, the rare exceptions exist and matter more than the rule for anyone treating this as evidence. The Bombay phenotype (homozygous null at the H locus) makes a person type as group O even when they carry an A or B allele, because there is no H precursor for the A or B enzyme to modify; such a parent can pass on an A allele and have a group-A child. Weak ABO subgroups can also mistype as O. Neither is common, but both are documented, which is why blood typing is never used on its own to establish parentage.
Why does the calculator ask for genotypes instead of blood types?
Because a blood type does not determine a genotype, and pretending otherwise would mean inventing numbers. Group A is genotype AA or AO, group B is BB or BO, and RhD positive is DD or Dd — routine serology cannot distinguish these. To turn 'group A' into probabilities a calculator would have to assume how often group-A people are AA rather than AO, which depends on ABO allele frequencies that differ substantially between populations. That assumption would be invisible in the output and would make an exact-looking number into a guess. Asking for the genotype keeps the arithmetic deterministic. If you do not know the genotype, run each compatible option and take the union of the possible types: that set is correct regardless of allele frequencies, even though the probabilities are not determined.
How is the Rh factor inherited, and can two Rh-positive parents have an Rh-negative child?
Yes, easily — if both are heterozygous. RhD positivity comes from having at least one working copy of the RHD gene on chromosome 1p34–36; RhD negativity is the absence of a working copy, most often an outright deletion of RHD in European populations and more often an RHD pseudogene or a hybrid RHD–RHCE gene in African populations. Two Dd parents are both RhD positive, and one quarter of their children inherit no working copy from either and are RhD negative, which is what the calculator returns as 25 %. The 'd' used here is a notational convenience, not a gene product: there is no competing 'd' protein, just nothing where RHD would be. This page models D only. Weak D and partial D variants — quantitative and qualitative changes to the antigen — type inconsistently between reagents and are outside the model.
Can this tell me who a child's father is?
No. Blood groups can occasionally exclude a claimed parent and can never confirm one, and even the exclusions have documented exceptions. The strongest classical exclusions — an AB parent cannot have a group-O child, two group-O parents can only have group-O children — are broken respectively by the cis-AB allele and by the Bombay phenotype. Cis-AB carries a single enzyme with both A and B activity, so both antigens come from one allele; Chun and colleagues (Annals of Laboratory Medicine, 2019) describe 'the birth of an O child from an AB mother' for exactly this reason, and report the allele at 0.0354 % of Koreans, 0.0012 % of Japanese and 0.00066 % of Chinese. Add weak D, partial D, ABO subgroups, chimerism and the effect of recent transfusion, and the picture is clear: parentage questions require accredited DNA testing. This page is a genetics teaching tool.
What does codominance mean for the AB blood group?
It means both alleles are fully expressed rather than one masking the other or the two blending. The A allele encodes a glycosyltransferase that adds one terminal sugar to the H antigen on the red cell; the B allele encodes a different transferase that adds a different sugar. Someone with genotype AB makes both enzymes, so both sugars appear on their red cells and a laboratory can detect each independently — this is not an intermediate 'half-A half-B' antigen, which is what incomplete dominance would produce. MedlinePlus Genetics defines codominance as two alleles both being expressed with each making a slightly different protein, and ABO is the standard human example. The O allele, by contrast, is recessive because it makes no working enzyme at all, so there is nothing for it to contribute in a heterozygote.
Why does the most-likely output sometimes list several blood types?
Because they are genuinely tied and picking one would be misleading. In the default example — AO × BO with both parents Dd — all four ABO groups come out at 25 % and RhD positive at 75 %, so O positive, A positive, B positive and AB positive all sit at exactly 18.75 %. Reporting only one of them would imply a distinction the genetics does not support. The calculator therefore lists every combination that achieves the maximum. The same logic applies to the excluded-groups output: it lists all groups at exactly zero, which for many parent pairs is the single most informative thing on the page.
Do ABO and Rh results really multiply?
Yes, because the two genes are on different chromosomes. ABO is at 9q34 and RHD at 1p34–36, so they assort independently in meiosis and knowing a child's ABO group tells you nothing about their RhD status. That makes the joint probability the simple product of the two: P(A positive) = P(group A) × P(RhD positive), which for the worked example is 25 % × 75 % = 18.75 %. Independence would fail for genes close together on the same chromosome — for instance RHD and RHCE are so tightly linked that they are effectively inherited as a haplotype — but that does not apply across ABO and RHD.

References& sources.

  1. [1]Dean L. (2012, updated 2015). ABO Blood Group. In: MINI Medical Genetics Summaries, National Center for Biotechnology Information (US), Bookshelf ID NBK100894. Source for the dominance convention used throughout this page: 'The ABO blood type is inherited in an autosomal codominant fashion. The A and B alleles are codominant, and the O allele is recessive.' Free full text. Retrieved 2026-07-29.
  2. [2]Dean L. (2005). Blood Groups and Red Cell Antigens, chapter 5 'The ABO blood group'. NCBI (US), Bookshelf ID NBK2267. Source for the molecular basis: the ABO locus has three allelic forms A, B and O; the A and B alleles encode glycosyltransferases producing the A and B antigens while 'the O allele encodes an enzyme with no function'; and the Bombay phenotype, in which individuals homozygous for null H alleles 'cannot produce A and B antigens'. Free full text. Retrieved 2026-07-29.
  3. [3]Dean L. (2005). Blood Groups and Red Cell Antigens, chapter 7 'The Rh blood group'. NCBI (US), Bookshelf ID NBK2269. Source for the RHD/RHCE gene pair, and for the molecular causes of the RhD-negative phenotype differing by ancestry — RHD deletion in European populations, versus RHD deletion, an RHD pseudogene or an RHD hybrid gene in African populations. The population percentages this chapter quotes are context only; no frequency is used in any calculation on this page. Free full text. Retrieved 2026-07-29.
  4. [4]Rosenkrans D., Zubair M. & Doyal A. Rh Blood Group System. StatPearls, NCBI Bookshelf ID NBK594252, last update 2 August 2023. Independent confirmation of the Rh model used here: RHD and RHCE at chromosome 1p34–36, 'the deletion of RHD or mutations resulting in a premature stop codon produces the D-negative phenotype', the autosomal dominant inheritance of the Rh factor, and the definitions of weak D ('weak or absent RBC agglutination by anti-D antibodies during routine serologic testing') and partial D ('a qualitative change to the D antigen epitope'). Free full text. Retrieved 2026-07-29.
  5. [5]Chun S., Choi S., Yu H. & Cho D. (2019). Cis-AB, the Blood Group of Many Faces, Is a Conundrum to the Novice Eye. Annals of Laboratory Medicine 39(2):115–120. doi:10.3343/alm.2019.39.2.115, PMID 30430772. Peer-reviewed source for the documented exception to the classical model: 'Both A and B antigens are expressed from the same allele inherited from a single parent', producing cases such as 'the birth of an O child from an AB mother'; typical phenotype A2B3; frequency 0.0354 % in Korea, 0.0012 % in Japanese and 0.00066 % in Chinese populations. Open access via PMC. Retrieved 2026-07-29.
  6. [6]MedlinePlus Genetics, US National Library of Medicine. Inheritance Patterns, page last updated 19 April 2021. Source for the definition of codominant inheritance: 'Two different versions (alleles) of a gene are expressed, and each version makes a slightly different protein. Both alleles influence the genetic trait.' Free. Retrieved 2026-07-29.
Related calculators
5
Punnett Square CalculatorgenSex-Linked Inheritance CalculatorgenPedigree Probability CalculatorgenProbability CalculatorstaPercentage Calculatorper

In this category

Embed

Quanta Pro

Paid features are coming later.

  • All 682 calculators remain free
  • No billing is enabled
Coming soon