Parents' Blood Types and the Child's Blood Type: Chances for All 10 Combinations, Impossible Types, and How Inheritance Works
"We're both type A, so how is our child type O?" "Can a parent with type AB have a type O child?" Blood type inheritance is simple: you carry two of the three genes A, B and O, one from each parent. The catch is that you can't tell from a blood test whether a type A person is AA or AO, so the answer has to be worked out as a probability. This guide covers the chance of each blood type for all 10 parent combinations, a chart of all 21 genotype pairings, how to estimate whether someone is AA or AO from population data, narrowing it down with the grandparents' blood types, which types are impossible when you work backwards from a child, how Rh negative is inherited, and exceptions such as cis-AB, with the calculation shown at each step.
The short answer: what each of the 10 parent combinations can and can't produce
If you don't care which parent is which, there are 10 combinations of parents' blood types. Only type O + type O fixes the child's blood type; every other pairing can produce two to four different types. Estimating each parent's genotype from the US blood type distribution (type O 45%, type A 40%, type B 11%, type AB 4%), the chances for the child are as follows.
| Parents | Type A | Type B | Type O | Type AB | Impossible for the child |
|---|---|---|---|---|---|
| A × A | about 82.3% | 0% | about 17.7% | 0% | B, AB |
| A × B | about 27.4% | about 22.2% | about 19.9% | about 30.5% | None (all four are possible) |
| A × O | about 57.9% | 0% | about 42.1% | 0% | B, AB |
| A × AB | 50% | about 21.1% | 0% | about 28.9% | O |
| B × B | 0% | about 77.7% | about 22.3% | 0% | A, AB |
| B × O | 0% | about 52.7% | about 47.3% | 0% | A, AB |
| B × AB | about 23.6% | 50% | 0% | about 26.4% | O |
| O × O | 0% | 0% | 100% | 0% | A, B, AB |
| O × AB | 50% | 50% | 0% | 0% | O, AB |
| AB × AB | 25% | 25% | 0% | 50% | O |
The "impossible" column doesn't mean unlikely; it means the chance is zero under normal inheritance. A parent with type AB never has a type O child, and two type O parents only have type O children. The figures marked "about" are averages that depend on whether a type A parent is AA or AO (or a type B parent BB or BO), estimated from population data. If both type A parents are actually AO, the chance of a type O child is 25%; if either one is AA, it is 0%.
How blood type is inherited: three genes, six genotypes
The ABO blood group is controlled by a gene with three versions (alleles): A, B and O. Everyone carries two, one inherited from the father and one from the mother. Which of their two a parent passes on is a 50/50 chance, so a child ends up with one of the father's two and one of the mother's two.
The blood type you see on a test (the phenotype) relates to the pair of genes you carry (the genotype) as shown below. A and B are dominant over O (O stays hidden when paired with them), and A and B are codominant (both show up together).
| Blood type | Genotype | Gene passed to a child | Estimated share of the US population |
|---|---|---|---|
| Type A | AA | Always A | about 6.3% |
| Type A | AO | A or O, half the time each | about 33.7% |
| Type B | BB | Always B | about 0.6% |
| Type B | BO | B or O, half the time each | about 10.4% |
| Type O | OO | Always O | about 45.0% |
| Type AB | AB | A or B, half the time each | about 3.9% |
This table explains most of the combinations at once. A type AB parent has no O to pass on, so they can't have a type O (OO) child. A type O parent can only pass on O, so two type O parents always have a type O child. And two type A parents have a type O child only when both are AO, carrying a hidden O, and both happen to pass it on.
Working out the odds with a Punnett square
Once you know both parents' genotypes, a 2×2 grid called a Punnett square gives you the child's odds. Write the father's two genes down the side and the mother's two across the top, then fill in each cell where they meet. Each of the four cells is equally likely, a 1 in 4 chance.
- 1
Write down the parents' genotypes
For example, take an AO father (type A) and a BO mother (type B). Put A and O down the side for the father and B and O across the top for the mother.
- 2
Fill in the four cells
A meets B to give AB, A and O give AO, O and B give BO, and O and O give OO. Order doesn't matter: AO and OA are the same genotype.
- 3
Convert to blood types and count
AB is type AB, AO is type A, BO is type B and OO is type O, so each of the four blood types fills one cell: 25% each. This pairing is why A × B parents can have a child of any blood type.
| Father AO × Mother BO | B from the mother | O from the mother |
|---|---|---|
| A from the father | AB (type AB) | AO (type A) |
| O from the father | BO (type B) | OO (type O) |
With the same A × B blood types, an AA father and a BB mother fill all four cells with AB, so the child is always type AB. An AO father and a BB mother give two cells of AB and two of BO, so type AB and type B are 50/50. That is why blood types alone can't pin down the odds: the same blood type can hide different genotypes.
Chart of all 21 genotype combinations
There are 21 pairings of the six genotypes if you don't distinguish father from mother. When the genotypes are known, every chance in the chart is 0, 1/4, 1/2, 3/4 or 1.
| Parents' genotypes | Type A | Type B | Type O | Type AB |
|---|---|---|---|---|
| AA × AA | 1 | 0 | 0 | 0 |
| AA × AO | 1 | 0 | 0 | 0 |
| AA × BB | 0 | 0 | 0 | 1 |
| AA × BO | 1/2 | 0 | 0 | 1/2 |
| AA × OO | 1 | 0 | 0 | 0 |
| AA × AB | 1/2 | 0 | 0 | 1/2 |
| AO × AO | 3/4 | 0 | 1/4 | 0 |
| AO × BB | 0 | 1/2 | 0 | 1/2 |
| AO × BO | 1/4 | 1/4 | 1/4 | 1/4 |
| AO × OO | 1/2 | 0 | 1/2 | 0 |
| AO × AB | 1/2 | 1/4 | 0 | 1/4 |
| BB × BB | 0 | 1 | 0 | 0 |
| BB × BO | 0 | 1 | 0 | 0 |
| BB × OO | 0 | 1 | 0 | 0 |
| BB × AB | 0 | 1/2 | 0 | 1/2 |
| BO × BO | 0 | 3/4 | 1/4 | 0 |
| BO × OO | 0 | 1/2 | 1/2 | 0 |
| BO × AB | 1/4 | 1/2 | 0 | 1/4 |
| OO × OO | 0 | 0 | 1 | 0 |
| OO × AB | 1/2 | 1/2 | 0 | 0 |
| AB × AB | 1/4 | 1/4 | 0 | 1/2 |
A type O child is only possible when both parents carry an O (AO, BO or OO on both sides). Only six rows have a non-zero type O column; if AA, BB or AB appears on either side, a type O child can't happen.
Is a type A person AA or AO? Estimating it from population data
A standard blood test reports both AA and AO simply as "type A". Confirming the genotype takes a genetic test, but if you know a population's blood type distribution, you can estimate what share of type A people are AO. The tool for this is the Hardy–Weinberg principle from high school biology.
- 1
Find the frequency of the O gene
Type O people are OO, so the share of type O equals the O frequency squared. In the US, type O is 45%, so the O frequency is √0.45 ≈ 0.671 (67.1%).
- 2
Find the A and B frequencies
People who are type A or type O carry only A and O, so their combined share equals (A + O) squared. √(0.40 + 0.45) − 0.671 ≈ 0.251 is the A frequency; B works out the same way at about 0.077. (This is known as Bernstein's method.)
- 3
Turn them into genotype shares
AA is A × A = 0.251² ≈ 6.3%, and AO is 2 × A × O = 2 × 0.251 × 0.671 ≈ 33.7%. Of all type A people (6.3% + 33.7%), about 15.8% are AA and about 84.2% are AO.
| Among people with this blood type | Japan | South Korea | China (Han) | United States |
|---|---|---|---|---|
| Type A who are AA | about 20.3% | about 19.6% | about 16.1% | about 15.8% |
| Type A who are AO | about 79.7% | about 80.4% | about 83.9% | about 84.2% |
| Type B who are BB | about 13.4% | about 16.7% | about 14.9% | about 5.5% |
| Type B who are BO | about 86.6% | about 83.3% | about 85.1% | about 94.5% |
In every one of these countries, more than 80% of type A and type B people carry an O. Because O is so common, few people end up with two copies of A or B. These estimates give the figures in the chart at the top: two type A parents have a type O child only when both are AO, so 0.842 × 0.842 × 1/4 ≈ 17.7%. For A × O, the type A parent has to be AO and pass on the O: 0.842 × 1/2 ≈ 42.1%.
Narrowing it down with the grandparents (why blood type seems to skip a generation)
A parent's genotype can be narrowed down from their own parents' blood types (the child's grandparents). If either parent of a type A mother is type O, that grandparent could only pass on O, so the mother must carry one O and is AO. Working through the possibilities for a type A mother gives the following (estimated with US data).
| Parents of the type A mother | Mother is AA | Mother is AO | Why |
|---|---|---|---|
| Either one is type O | 0% | 100% | A type O parent can only pass on O |
| Type A × type B | 0% | 100% | Getting B from the type B parent would make her type AB, so she got O |
| Type A × type A | about 41% | about 59% | Both AA and AO are possible |
| Type AB × type AB | 100% | 0% | Neither parent has an O |
| Type B × type B | — | — | Neither parent has an A, so a type A child is impossible |
Once the mother is known to be AO, the chance of a type O child with a type O father rises from the estimated 42.1% to 50%. If instead her parents are both type A, the chance she is AA rises to about 41%, and the chance of a type O child drops to about 29.6%.
"Both parents are type A, but the child is type O like grandpa" is often described as blood type skipping a generation, but it's ordinary inheritance. The type A parents were carrying, without showing it, an O inherited from the type O grandfather, and in the child's generation two O's happened to meet.
Working backwards from the child: which parent blood types are possible or impossible
When you know the child's blood type and one parent's, the possible and impossible blood types for the other parent are as follows. Each percentage is "the chance a person has that blood type × the chance that pairing produces the child's type", scaled to add up to 100% (US data).
| Child | Known parent | Possible types for the other parent | Impossible types |
|---|---|---|---|
| Type O | Type A, B or O | O about 67%, A about 25%, B about 8% | AB |
| Type O | Type AB | — (this combination itself is impossible) | All |
| Type A | Type A | A about 51%, O about 41%, B about 5%, AB about 3% | None |
| Type A | Type B or O | A about 92%, AB about 8% | B, O |
| Type A | Type AB | O about 49%, A about 43%, B about 6%, AB about 2% | None |
| Type B | Type B | O about 55%, A about 21%, B about 20%, AB about 5% | None |
| Type B | Type A or O | B about 75%, AB about 25% | A, O |
| Type B | Type AB | O about 60%, A about 23%, B about 15%, AB about 3% | None |
| Type AB | Type A | B about 75%, AB about 25% | A, O |
| Type AB | Type B | A about 92%, AB about 8% | B, O |
| Type AB | Type AB | A about 70%, B about 18%, AB about 12% | O |
| Type AB | Type O | — (this combination itself is impossible) | All |
For a type O child, the estimate for the other parent (O 67%, A 25%, B 8%) is the same whether the known parent is type A, B or O, and it matches the US frequencies of the O, A and B genes exactly. Think about who passed the child an O: a type O person always does, and an AO or BO person does half the time, which brings the result back to the gene frequencies.
Rh positive and Rh negative: two Rh+ parents can have an Rh− child
The Rh blood type is controlled by a separate gene from ABO. It has two versions, D (Rh+) and d (Rh−). D is dominant, so DD and Dd are Rh+, and only dd is Rh−. As with ABO, one copy comes from each parent.
About 15% of people in the US are Rh−, which puts the frequency of the d gene at √0.15 ≈ 38.7%. That means about 55.8% of Rh+ people, more than half, are estimated to be Dd, carrying a hidden d. If both Rh+ parents are Dd, each child has a 1 in 4 chance of being Rh−.
| Parents' Rh | Chance the child is Rh− (US) | Calculated with Japan's data |
|---|---|---|
| Rh+ × Rh+ | about 7.8% | about 0.4% |
| Rh+ × Rh− | about 27.9% | about 6.6% |
| Rh− × Rh− | 100% | 100% |
ABO and Rh are inherited independently, so the chance of a combination like "O negative" is the product of the two separate chances. For A+ and O+ parents with US data, that's a type O chance of about 42.1% × an Rh− chance of about 7.8% ≈ 3.3%. Also, an Rh+ person with an Rh− parent must have inherited a d, so they are definitely Dd.
When the chart doesn't apply: cis-AB, the Bombay phenotype, subtypes and newborn tests
All of the calculations above assume the A, B and O genes are passed on by the textbook rules. In rare cases like these, parents and children end up with combinations the chart says are impossible.
- Cis-AB: a rare variant where a single gene does the work of both A and B. A cis-AB person can pass on "AB" or "O", so AB + O parents can have a type AB child, and a type AB parent can have a type O child. It has been reported most often in East Asia, especially Korea and Japan.
- Bombay phenotype: an extremely rare type in which the body can't make the H substance that A and B markers (antigens) are built on. Even with A or B genes, the person tests as type O, so it can look as if a "type O parent" had a type A or type B child. It's named after Mumbai (formerly Bombay), India, where it was first found.
- Subtypes (A2, Am, B3 and others): the A or B markers are weak, so a type A person may test as type O, or the result may be hard to call.
- Bone marrow (stem cell) transplant: if the donor has a different blood type, the recipient's blood type changes to the donor's. The genes they pass to their children don't change, though.
- Newborn tests: a newborn doesn't yet have enough of the antibodies (anti-A and anti-B) in their blood, so results can change or be left undetermined. Many hospitals don't routinely type a baby's blood at birth at all.
If your family's blood types don't fit the chart, first consider these exceptions, a misremembered blood type, or an old test result. You can get your blood type rechecked when donating blood or before surgery. When parentage actually needs to be established, a DNA test is used rather than blood type.
Do the odds change by country? Japan, South Korea, China and the US compared
The figures marked "about" shift a little depending on which population's blood type distribution you use for the estimate. With data for Japan, South Korea, China (Han) and the US, they come out like this.
| Japan | South Korea | China (Han) | United States | |
|---|---|---|---|---|
| Blood type shares (A, O, B, AB) | 38.6, 30.3, 21.6, 9.5% | 34, 28, 27, 11% | 31, 34, 28, 7% | 40, 45, 11, 4% |
| A × A → type O | about 15.9% | about 16.2% | about 17.6% | about 17.7% |
| B × B → type O | about 18.8% | about 17.3% | about 18.1% | about 22.3% |
| A × O → type O | about 39.9% | about 40.2% | about 42.0% | about 42.1% |
| A × B → type AB | about 34.1% | about 34.9% | about 33.3% | about 30.5% |
| Share who are Rh− | about 0.5% | about 0.3% | about 0.4% | about 15% |
| Rh+ × Rh+ → Rh− | about 0.4% | about 0.3% | about 0.4% | about 7.8% |
ABO odds differ by only a few points from country to country, but Rh odds differ hugely. With about 15% of people Rh− in the US, more than half of Rh+ people are estimated to carry a d, so two Rh+ parents are about 18 times as likely to have an Rh− child as in Japan. If the parents come from different places, it makes sense to use each parent's own population.
Run the numbers for your own family with the blood type calculator
Every calculation in this guide can be run for your own family with the Blood Type Inheritance Calculator. It works entirely in your browser, and the blood types you enter are never sent to a server.
- 1
Choose the father's and mother's blood types
Under "Child's blood type from parents", pick type A, B, O, AB or Unknown for each parent to see the chance of each blood type for the child and which types are impossible.
- 2
Add genotypes or grandparents if you know them
If you know a genotype such as AA or AO, you can select it instead. Enter the grandparents' blood types and the calculator narrows down the parents' genotypes as in the table above and recalculates.
- 3
Choose Rh and the population
Pick Rh+ or Rh− to get the chance of combinations like "A positive". The blood type distribution used for the estimates can be set to Japan, South Korea, China (Han) or the United States.
- 4
Work backwards from a child
Under "Parent's blood type from a child", choose the child's type and the known parent's type to see which types are possible and impossible for the other parent.
The possible genotype pairings are also shown as a Punnett square, so it works for checking Mendel's laws in a biology class too. For counting the days to a new baby's milestones and checkups, see how to count a baby's days since birth.
FAQ
- Can two type A parents have a type O child?
- Yes. If both parents are AO, there is a 1 in 4 chance the child is type O (OO). Based on US blood type data, about 84% of type A people are AO, so two type A parents have about a 17.7% chance of a type O child. They can't have a type B or type AB child.
- What blood types can type A and type B parents have?
- Any of the four. Estimated with US data, the chances are about 30.5% type AB, 27.4% type A, 22.2% type B and 19.9% type O. If the parents are AO and BO, each type is 25%; if they are AA and BB, the child is always type AB.
- Can type AB and type O parents have a type AB or type O child?
- Not under normal inheritance: the child is type A or type B, 50% each. The AB parent passes on A or B and the O parent passes on O, so the child is AO (type A) or BO (type B). A parent with the rare cis-AB variant, however, can have a type AB or type O child.
- Which blood type is impossible for a parent of a type O child?
- Type AB. A type O child gets an O from each parent, so a type AB parent, who has no O, can't have one (except with rare variants like cis-AB). If one parent is type A, B or O, US data puts the other parent at about 67% type O, 25% type A and 8% type B.
- How can I find out whether I'm AA or AO?
- A standard blood test can't tell; only a genetic test can say for sure. But if either of your parents is type O, or any of your children is type O, you must be AO. A type A person whose parents are type A and type B is also AO. Otherwise, US data puts a type A person at about 16% AA and 84% AO.
- Can two Rh positive parents have an Rh negative baby?
- Yes. If both parents are Dd (Rh+ but carrying the Rh− gene), each child has a 1 in 4 chance of being Rh−. In the US, more than half of Rh+ people are estimated to be Dd, so two Rh+ parents have about a 7.8% chance of an Rh− child, and an Rh+ and an Rh− parent about 27.9%.
- If a child has a grandparent's blood type, did it skip a generation?
- It's ordinary inheritance. For example, a type A parent (AO) who inherited an O from a type O grandfather can pass that O on; if the other parent also passes on an O, the child is type O like the grandfather. The O was there in the parent's generation, just not showing.
- Are there parents and children whose blood types don't fit the chart?
- Rarely, yes: cis-AB, the Bombay phenotype that tests as type O, weak subtypes, and people whose blood type changed after a bone marrow transplant. Tests done in infancy can also change, and misremembered blood types are common. Don't judge parentage from blood type alone; get a retest if something seems off.
- Do a child's blood type odds differ by country?
- Slightly, because the chance of being AA or AO depends on a country's blood type mix. Two type A parents have a type O child about 15.9% of the time with Japan's data, 16.2% with South Korea's, 17.6% with China's and 17.7% with US data. Rh differs a lot more: two Rh+ parents have an Rh− child about 7.8% of the time in the US but about 0.4% in Japan.
Tools in this article
Related articles
Normal Resting Heart Rate: Averages by Age and Sex, Causes of a High or Low Pulse, and How to Measure It Correctly
Your smartwatch or a checkup says "resting heart rate 72", but is that normal, high or low? The normal adult range is usually given as 60–100 beats per minute, yet 60 and 95 say quite different things about your body, and children and endurance athletes have different ranges altogether. This guide sorts the average and normal resting heart rate by age (infants, toddlers, school-age children, adults), by sex and by fitness level, then covers how to measure it properly first thing in the morning and the error built into a 15-second count, how a wrist count, a phone camera and a smartwatch differ, what can make it high or low and when to see a doctor, how to lower it, and how to use it to set Karvonen target heart rates. The age ranges and heart rate zone figures use the same reference values and formulas as the Heart Rate Monitor tool.
Read article →How to Calculate Projector Throw Distance: How Far for 100 Inches, Reading the Throw Ratio, Screen Size for Your Room and Mounting Height
The most common projector regret is finding out after setup that the picture is smaller than expected, or that a 100-inch image runs off the top of the wall. Both can be checked before you buy, using the throw ratio and the screen dimensions. This guide starts from the basic formula, throw distance = throw ratio × image width, and works through the distance you need for 100 and 120 inches, how to work backwards from the distance you have to the screen size you get, charts for standard, short throw and ultra short throw projectors, how lens height and offset decide where the image sits on a table or ceiling mount, the largest size that fits a small or medium room, viewing distance guidelines, and whether your lumens are enough. All figures are calculated with the same formulas as the Projector Throw Distance Calculator.
Read article →Darts Checkout Chart and How to Remember It: 01 Finishes for Open Out and Double Out, Plus Bogey Numbers
You are playing 01, your score drops below 100, and suddenly you freeze: where do I aim to finish from here? The right checkout depends on two settings: whether the last dart has to land in a particular ring (open out, double out or master out), and whether the bull is a single 50 or split into 25 and 50 (fat bull or split bull). This guide covers the highest checkout and the bogey numbers you cannot finish in three darts for each rule, checkout charts by remaining score, why D16 and D20 are the doubles to leave, how to fix an odd number, the one dart to throw when you cannot finish this turn, an easy way to memorise checkouts, and how to read a rough rating from PPD, PPR and count-up scores. Every number in the tables was checked by enumerating all 1 to 3-dart combinations with the same logic as the Darts Checkout Chart tool.
Read article →