Most online blood type cross answers are wrong in a specific, repeatable way. Ask what a type A and a type O parent can produce and a lot of sources will tell you 50% type A and 50% type O. That is only correct if you already know the type A parent is heterozygous. A routine blood test cannot tell AA from AO, so if all you are given is a phenotype, the honest answer has to average over both possibilities — and it lands somewhere else entirely.
This calculator does the averaging properly and shows you the percentages that come out of it, including the awkward ones.
Ask for the offspring of a type A and a type O parent and most answers online will tell you 50/50. That is only true if you already know the A parent is heterozygous. A blood test cannot tell AA from AO, so the honest answer averages over both and lands somewhere else. This averages properly and shows the working.
Type A is either AA or AO, and type B is either BB or BO. A and B are both dominant over O, and A and B are dominant over each other, which is codominance. Enter known genotypes if your question gives them and the answer is exact instead of averaged.
The ABO system in one minute
ABO blood type is controlled by three alleles: IA, IB and i. IA and IB are codominant, so a person with one of each has type AB. Both are dominant over i, which is recessive.
- Type A is AA or AO
- Type B is BB or BO
- Type AB is always AB
- Type O is always OO
AB and O are the only two phenotypes that pin down a genotype exactly. A and B each hide two possibilities, which is where most of the confusion comes from.
How to work out a cross
Each parent passes on one of their two alleles, with equal probability. So for a known cross you can draw a Punnett square directly. For example, AB × O: the AB parent passes either IA or IB, and the O parent can only pass i. Every child is therefore either IAi (type A) or IBi (type B), giving 50% A and 50% B — never AB, and never O.
Now the same idea with a hidden genotype. A type A parent is AA half the time and AO half the time:
- If they are AA, every child with an O parent is type A.
- If they are AO, half the children are type A and half are type O.
Average those: 0.5 × 1 + 0.5 × 0.5 = 0.75, so 75% type A and 25% type O. Not 50/50.
Type A × type B is the one most people get wrong. Averaging over AA/AO and BB/BO gives 18.75% A, 18.75% B, 56.25% AB and 6.25% O. The tempting 25% each answer comes from quietly assuming both parents are heterozygous, which the phenotypes never told you.
Why this matters beyond the exam
A blood test detects the antigens on the red cell surface, not the alleles. Anyone with an A or B phenotype could be homozygous or heterozygous, and nobody finds out without a family history or genetic testing. That is genuinely useful in transfusion medicine: a type O negative patient can only safely receive O negative red cells, and knowing a donor is OO rather than AO matters when you are matching across generations.
There is also one famous exception. The Bombay phenotype, hh, switches off the H antigen that A and B are built on. A person with hh has a completely normal A, B or AB genotype but tests as type O, because the test looks for the antigen and it is not there. It is rare, and it is the standard answer when a question asks why a blood test and a genotype can disagree.
Worked questions this calculator answers
- Type A × type O. Answer: 75% A, 25% O, averaged over AA and AO.
- Type AB × type O. Answer: 50% A, 50% B.
- Type A × type B. Answer: 18.75% A, 18.75% B, 56.25% AB, 6.25% O.
- Type AB × type AB. Answer: 25% A, 25% B, 50% AB, and never O.
- Which crosses give only one phenotype? Answer: O × O gives all O, and AB × AB is the only other one that fixes a genotype at both ends.
If your question hands you genotypes, use them rather than phenotypes and the answer is exact. If it hands you only phenotypes, the averaged percentages above are the correct ones, and the reason they look strange is the averaging itself, not a mistake in the arithmetic.
Frequently Asked Questions
What offspring does a type A and type O parent produce?
75% type A and 25% type O. The 50/50 answer you often see is only right if the type A parent is known to be heterozygous.
Why does A x O give 75% rather than 50%?
Because a blood test cannot distinguish AA from AO. Half the time the type A parent is AA and every child is type A; half the time it is AO and only half the children are. Averaging gives 0.75.
Can an AB parent and an O parent have a type B child?
Yes. The AB parent passes on either Iᴬ or Iᴮ, and the O parent always passes on i, so children are Iᴬi (type A) or Iᴮi (type B) in equal measure.
What does type A x type B give?
18.75% A, 18.75% B, 56.25% AB and 6.25% O, once you average over the genotypes each phenotype hides.
What is codominance in the ABO system?
Iᴬ and Iᴮ are both expressed in a heterozygote, so someone with one of each has type AB rather than an intermediate type.
What is the Bombay phenotype?
A rare genotype, hh, that switches off the H antigen that A and B are built on. The person can carry a normal A, B or AB genotype but tests as type O.
Why is a blood test not the same as a genotype?
A blood test detects antigens on the red cell surface, not the alleles behind them. Type A could be AA or AO, and nothing routine will tell you which.
How do I do a Punnett square for blood type?
This is the one trait where a Punnett square genuinely belongs, because ABO is a single gene, so the square is always a 2×2. Write the gametes from each parent across the top and down the side, then fill the four boxes. A type A parent passes on either IA or i, and a type O parent can only pass on i, which gives three boxes of IAi and one of ii, so three type A to one type O. Type AB is the one to watch, because IA and IB are codominant rather than one dominating the other, so a box containing both reads as type AB instead of picking a winner.