How to Use Punnett Squares to Predict Genetic Traits
By the BrainSnail editorial team. How these articles are written and checked, and how to tell us when one is wrong.
Using Punnett squares becomes much easier when you treat them as probability grids, not as pictures of exactly what a family will look like. Each square shows one possible allele combination an offspring could inherit from two parents.
Filling in the square, step by step
A Punnett square starts with the alleles each parent can pass on. Suppose a gene has two versions, B and b, and both parents have the genotype Bb. Each parent can place either B or b into a sex cell, so write B and b across the top of the grid for one parent and B and b down the side for the other. Then combine the row and column letters inside the four boxes. The possible offspring genotypes are BB, Bb, Bb, and bb.
This is the core of using Punnett squares. The grid does not predict the exact order of children or promise that four children will include one of each box. It shows probabilities for each independent fertilisation event. In this example, there is a 25 percent chance of BB, a 50 percent chance of Bb, and a 25 percent chance of bb for each offspring.
Keep genotype and phenotype separate. Genotype is the allele combination, such as Bb. Phenotype is the observable trait produced by that genotype together with biological context. In a simple complete-dominance example, B may produce the dominant phenotype in both BB and Bb, while bb produces the recessive phenotype.
Read dominant and recessive alleles carefully
When using Punnett squares, many mistakes come from assuming that a dominant allele is stronger, better, or more common. Dominant only describes how an allele is expressed when two different alleles are present. A recessive allele can be common in a population, and a dominant allele can be rare. The capital and lowercase letters are simply a notation system.
A heterozygous individual has two different alleles, such as Bb. A homozygous individual has two matching alleles, such as BB or bb. If B is completely dominant over b, both BB and Bb show the dominant phenotype. This means a person or organism can carry a recessive allele without showing the recessive trait.
Real genetics can be more complicated than the simplest Punnett square. Some traits involve incomplete dominance, codominance, several genes, environmental effects, or sex-linked inheritance. The grid is still useful, but the rules for translating genotype into phenotype may change.
Use probabilities instead of family predictions
A good way to check a Punnett square is to turn the boxes into fractions or percentages. Four equally likely boxes make the arithmetic simple. One box out of four is 25 percent, two boxes are 50 percent, and three boxes are 75 percent. If two boxes contain the same genotype, combine them before reporting the result.
Imagine a cross between Bb and bb. The first parent can pass B or b, while the second can pass only b. The four boxes become Bb, bb, Bb, and bb. The genotype probability is therefore 50 percent Bb and 50 percent bb. Under complete dominance, that also gives a 50 percent chance of the dominant phenotype and a 50 percent chance of the recessive phenotype.
Do not read a 25 percent result as a schedule. Four births could all show the same outcome because probability describes long-run chances, not a guaranteed pattern in a tiny sample. When a question asks you to explain the result, state the parent genotypes, the possible gametes, the offspring combinations, and then the probability.
Two genes at once and the product rule
Punnett squares can track two genes together, as long as the genes are inherited independently, which is true for genes on different chromosomes. Imagine pea plants where R gives round seeds and r wrinkled seeds, while Y gives yellow seeds and y green. A plant with genotype RrYy can make four kinds of gamete: RY, Ry, rY and ry. Crossing two such plants needs a grid of four by four, giving 16 boxes. Counting phenotypes gives the classic ratio of 9 round yellow, 3 round green, 3 wrinkled yellow and 1 wrinkled green.
You do not have to draw all 16 boxes every time. Because the genes are independent, you can treat each one as its own small square and multiply the probabilities. The chance of rr from Rr crossed with Rr is 1 in 4, and the chance of yy from Yy crossed with Yy is also 1 in 4, so the chance of a wrinkled green plant is one quarter times one quarter, or 1 in 16. This product rule is quicker and less error-prone, and it extends easily to three or more genes.
A worked example with a sex-linked gene
Some genes sit on the X chromosome, so the Punnett square has to track sex chromosomes as well as alleles. Females usually have two X chromosomes and males have one X and one Y, and the Y carries few of the same genes. Red-green colour blindness is a classic example, caused by a recessive allele on the X chromosome. Write the usual allele as XB and the colour-blind allele as Xb. Suppose a mother has normal vision but carries the allele, XBXb, and the father has normal vision, XBY. Her gametes carry XB or Xb, and his carry XB or Y.
The four boxes are XBXB, XBXb, XBY and XbY. Read them by sex. Both daughter boxes include the father's XB, so no daughter from this cross is expected to be colour-blind, although each has a one in two chance of being a carrier. Each son has a one in two chance of being XbY, because a son's single X comes from his mother and he has no second X to mask the recessive allele. This explains why red-green colour blindness is much more common in males, and why it can seem to skip a generation by passing quietly through carrier mothers.
A worked example with blood groups
The ABO blood groups show what happens when a gene has more than two alleles and some of them are codominant. Write the three alleles as IA, IB and i. IA and IB are codominant, so a person who inherits both shows both and has blood group AB. Each is dominant over i, so IAIA and IAi both give group A, while IBIB and IBi give group B. Only ii gives group O. A Punnett square still works in the usual way, but each parent places just two of the three possible alleles around the grid.
Now cross a parent with group A and genotype IAi with a parent with group B and genotype IBi. The four boxes are IAIB, IAi, IBi and ii, so each child has a one in four chance of being AB, A, B or O. Two parents with groups A and B can therefore have a child of any blood group. By contrast, two parents with group O are both ii and can normally pass on only i. Before DNA testing, blood groups were used in disputes about parentage, but they could only rule people out, never prove a match, and rare genetic variants meant even exclusions needed care.
The takeaway
Using Punnett squares comes down to four calm steps: identify the parent alleles, place possible gametes around the grid, combine them in the boxes, and convert the results into probabilities. Keep genotype separate from phenotype, and remember that the square predicts chances rather than the exact traits of future offspring.