Punnett squares predict ratios, not outcomes
The most common mistake with a Punnett square is treating a probability as a quota.
5 min read
What does a Punnett square show?
It shows the probability of each genotype among the offspring of a given cross, by laying out every combination of parental gametes. A monohybrid cross between two heterozygotes gives a 3:1 phenotypic ratio and a 1:2:1 genotypic ratio. Those are probabilities per offspring, not counts — each offspring is an independent event, so four offspring from that cross can easily be four of the same phenotype.
Probability, not prophecy
A 3:1 ratio means each offspring has a 3-in-4 chance of showing the dominant phenotype. It does not mean that in any four offspring, three will.
Each conception is independent, exactly like a coin toss. Four offspring all showing the recessive phenotype from a heterozygous cross is unlikely — about 1 in 256 — but entirely possible, and it happens.
The ratio describes the long run. It becomes reliable across hundreds of offspring, which is why genetics was worked out on pea plants and fruit flies rather than on families.
This is also why a couple who already have one affected child face exactly the same probability with the next.
Genotype and phenotype ratios differ
The classic monohybrid cross gives a 3:1 phenotypic ratio and a 1:2:1 genotypic ratio, and conflating them causes errors.
Three of the four boxes show the dominant trait, but they are not identical: one is homozygous dominant and two are heterozygous carriers. Those individuals look the same and behave very differently in the next generation.
For anything involving carriers — recessive conditions, breeding programmes — the genotype ratio is the one that matters, and it is invisible from the phenotype alone.
Our calculator reports both ratios and reduces them properly, including the test-cross case that produces a 1:1.
The test cross tells you what you cannot see
If an individual shows a dominant phenotype you cannot tell from looking whether it is homozygous or heterozygous. Crossing it with a homozygous recessive resolves it.
A homozygous dominant parent produces all dominant offspring. A heterozygous parent produces roughly half dominant, half recessive — a 1:1 ratio.
This is the classic method for determining an unknown genotype, and it remains the intuition behind a great deal of practical breeding even where genetic testing is now available.
Note that a single recessive offspring proves heterozygosity immediately, while a run of dominant offspring only makes homozygosity increasingly likely.
Where the simple model stops
Punnett squares assume one gene, two alleles, complete dominance and independent assortment. Plenty of real traits break at least one of those.
Incomplete dominance gives a blended heterozygote, so the phenotypic ratio becomes 1:2:1 rather than 3:1. Codominance expresses both alleles. Many traits involve multiple genes, producing continuous variation rather than discrete categories. Linked genes on the same chromosome do not assort independently at all.
Most human traits of interest — height, most disease risk — are polygenic, which is precisely why a Punnett square cannot predict them.
The square is a teaching tool for the simple case, and an excellent one. It is not a general model of inheritance.
For a real family question, see a genetic counsellor
If you are working out inheritance risk for an actual condition in an actual family, a Punnett square is not the right instrument.
Real risk assessment accounts for penetrance, variable expressivity, carrier frequency in the relevant population, family history across several generations, and often direct genetic testing. A counsellor does that properly and can explain what the numbers mean for your decisions.
In Canada, genetic counselling is generally accessed through a referral from your family doctor or specialist.
A calculator is useful for coursework and for understanding the mechanism. It is not a clinical assessment.