During sexual reproduction, a parent is equally likely to pass on to its offspring either of the two alleles it has at each genetic locus. This makes it possible to list and estimate the probability of specific genotypes being produced from the pairing of two individuals. Given two allele from each parent, four allele combinations are possible. These combinations and their probabilities can be readily visualized using a Punnett square.
To set up a single locus Punnett Square, the genotype of each parent is placed on the sides of a four chambered box. One parent’s alleles are placed across the top. The alleles of the other parent are placed down one side. The alleles on the edges guide how the central squares are filled in. Once complete, a Punnett square shows the genotypes possible from crossing two individuals. Each of the four boxes in the square contains one of the four possible genotypes. The genotype in each box has a 25% probability of occurring every time the two individuals are crossed. If two boxes contain the same genotype, the probability of that genotype occurring doubles to 50%.
Punnett squares are most commonly used to examine genotype probabilities from one genetic locus at a time. They can be used to look at more than one locus at time, but some find the resulting diagrams complicated and difficult to interpret.
The model below illustrates the use of a Punnett Square to determine the possible genotypes that can arise from mating two individuals with known genotypes. The organism in the model is a plant. The plant is diploid. The trait is flower color. Below the illustration is a youtube video demonstrating its use. There is also a problem set you can use to test your understanding of these concepts.