The Punnett square calculator above takes two parent genotypes, works out every gamete each parent can produce, pairs them exhaustively, and draws the resulting grid. It then counts the squares to give you a genotype ratio and a phenotype ratio. That is exactly the procedure done by hand on paper, with the difference that a dihybrid cross has sixteen squares and a trihybrid has sixty-four, and hand-counting sixty-four cells is where errors creep in rather than where understanding happens.
Arb Digital publishes this as part of a free science and maths tool library. Before going further, one thing needs saying plainly, because it governs everything below: this is a teaching tool for the classic Mendelian examples used in biology courses. It is not a predictor of any real person's traits, and it must never be used to estimate anybody's chance of inheriting a medical condition. Human characteristics almost never follow a single-gene pattern, and the ones that do are a matter for a clinical geneticist, not a web page.
What This Punnett Square Calculator Does
A Punnett square is a bookkeeping device invented to make the multiplication rule of probability visible. Each parent contributes one allele per gene to the offspring, chosen at random from the pair that parent carries. The square lists one parent's possible gametes along the top and the other's down the side, and each cell is one possible combination. Because every cell is equally likely, counting cells gives probabilities without any arithmetic at all — which is precisely why the device has survived since Reginald Punnett drew the first one.
The calculator handles one, two, or three genes. For each gene it splits the parent's two-letter genotype into its two alleles, then forms every combination across genes to build the gamete list: a parent with genotype AaBb produces AB, Ab, aB and ab. Offspring genotypes are written with the dominant allele first at each locus, so Aa and aA are recognised as the same genotype and counted together, which is a step that trips people up when tallying by hand.
Boundary worth stating: this tool answers questions about a specific cross between two known parents. If your question is about frequencies across a whole population rather than one mating, the Hardy-Weinberg calculator is the right page, and it works from allele frequencies rather than from parent genotypes.
How to Use It
- Pick the cross type. Selecting monohybrid, dihybrid or trihybrid loads a standard worked example so you can see the expected shape before entering your own data.
- Type each parent genotype as pairs of letters, uppercase for dominant and lowercase for recessive: Aa, AaBb, AaBbCc. Both parents must use the same gene letters in the same order.
- Read the phenotype ratio in the hero. For a dihybrid cross between two double heterozygotes this is the familiar 9:3:3:1.
- Read the genotype ratio in the note below the grid. It is always the more detailed of the two, because several genotypes can share one phenotype.
- Switch the grid to phenotype labels when you want to see visually which cells collapse into the same visible class.
The Formula and How It's Calculated
There is no equation as such, only a systematic expansion. A parent heterozygous at n genes produces 2n distinct gametes, so the grid for two such parents has 2n × 2n = 4n squares: four for a monohybrid cross, sixteen for a dihybrid, sixty-four for a trihybrid. Every square carries equal probability, which is 1 divided by the number of squares.
Each cell is assembled locus by locus. Take one allele from the row gamete and one from the column gamete for gene A, sort them so the uppercase allele comes first, and repeat for every gene. The tool then tallies identical genotype strings to build the genotype ratio, and separately maps each genotype to a phenotype class using complete dominance: a locus showing at least one uppercase allele expresses the dominant form, and only a locus with two lowercase alleles expresses the recessive one.
The classic dihybrid result of 9:3:3:1 falls straight out of this. Nine of sixteen squares are dominant at both loci, three are dominant at the first and recessive at the second, three are the reverse, and one is recessive at both. It is also exactly the product of two independent 3:1 ratios, which is Mendel's law of independent assortment restated as arithmetic: (3 + 1) × (3 + 1) expands to 9 + 3 + 3 + 1. The National Human Genome Research Institute's definition of dominant traits and alleles is the underlying rule the phenotype grouping relies on.
Why the Grid Is Slower Than Multiplying
For anything beyond two genes the square becomes an inefficient way to answer most questions. If you want the probability that an offspring of AaBbCc × AaBbCc shows all three dominant traits, drawing sixty-four cells and counting twenty-seven of them is a lot of work for an answer you can get in one line: each gene independently gives a 3/4 chance of the dominant phenotype, so the combined chance is 3/4 × 3/4 × 3/4 = 27/64. The forked-line method and the plain multiplication rule scale; the grid does not.
What the grid gives you in exchange is transparency. It shows why the numbers come out as they do, which matters when the question is not a simple product — asking for the probability of at least two dominant traits, for example, or of a specific genotype rather than a phenotype. Use the square to build intuition and to check awkward cases, and use multiplication once you trust the pattern. Our probability calculator handles the multiplication side when the crosses get large.
Where Simple Dominance Stops Working
This calculator assumes complete dominance, which is the standard assumption in introductory coursework and the one that produces the tidy 3:1 and 9:3:3:1 ratios. Real inheritance has several well-documented departures, and knowing them stops you misapplying a clean result.
Under incomplete dominance, the heterozygote shows an intermediate phenotype, so a monohybrid cross gives 1:2:1 for phenotype as well as genotype rather than 3:1. Under codominance, both alleles are expressed fully and simultaneously, as in the AB blood group. Multiple alleles break the two-letter notation entirely: the ABO locus has three common alleles, not two. Epistasis occurs when one gene masks another's expression, which distorts the 9:3:3:1 ratio into forms like 9:7 or 12:3:1. Linkage breaks independent assortment outright, because genes close together on the same chromosome tend to travel as a unit rather than sorting freely. And most traits people actually care about are polygenic, shaped by many genes plus environment, which no square of any size will model.
Why This Should Never Be Used on People
It is worth being blunt about this, because the search that brings people to a page like this is often a personal one. A Punnett square models a single gene with two alleles and complete dominance, in an organism where you know both parents' genotypes with certainty. Human eye colour, hair colour, height and skin tone are none of those things — eye colour alone involves multiple genes and does not follow the brown-dominant-over-blue rule taught in schools, and the rule was known to be an oversimplification long before genome sequencing confirmed it.
Inherited medical conditions are a separate and more serious matter. Even for conditions that genuinely are single-gene, carrier status, variable penetrance, new mutations and family-specific variants all mean that a grid on a web page tells you nothing reliable about a particular family. MedlinePlus Genetics, published by the US National Library of Medicine, sets out how the real inheritance patterns work and why professional genetic counselling exists. Use this tool for peas, fruit flies and coursework. For anything about a real person, see a clinician.
Reading a Ratio Correctly
A 3:1 ratio does not mean that four offspring will arrive as three dominant and one recessive. It means each individual offspring independently has a 3/4 chance of the dominant phenotype. Small families routinely depart from the expected split, exactly as four coin tosses routinely fail to give two heads and two tails, and there is no corrective mechanism that makes later offspring compensate for earlier ones. Ratios describe expectations over many events, not schedules.
This is also why real breeding data is tested statistically rather than compared by eye. If a cross expected to give 9:3:3:1 produces counts that look somewhat off, the question is whether the gap is larger than chance would comfortably produce, which is a goodness-of-fit problem — the same machinery our chi-square calculator applies. Converting the counted squares into percentages, which the percentage calculator makes quick, is usually the clearest way to present a result to somebody who does not think in ratios.
Arb Digital publishes hundreds of free, no-signup tools across science, statistics, finance and everyday maths. Nothing you type into any of them is stored or sold.
Browse All Free Tools Contact Arb DigitalCommon Mistakes to Avoid
- Writing gametes as genotypes. A gamete from AaBb is AB or Ab or aB or ab — one allele per gene, never two.
- Counting Aa and aA as different. They are the same genotype; the tool normalises them so tallies come out right.
- Confusing genotype and phenotype ratios. A monohybrid cross of two heterozygotes gives 1:2:1 genotypes but 3:1 phenotypes, and quoting the wrong one is the most common exam slip.
- Assuming independent assortment always holds. Genes on the same chromosome and close together are linked, and their offspring ratios depart from the grid's prediction.
- Applying the grid to human traits. Eye colour, height and almost every trait of personal interest are polygenic, and no Punnett square models them.
Related Free Tools From Arb Digital
For population-level allele frequencies rather than a single cross, use the Hardy-Weinberg calculator. The probability calculator handles the multiplication and addition rules that underpin every square, and the chi-square calculator tests observed breeding counts against expected ratios. For sequence work there is the DNA to mRNA converter. When you need to express a result differently, the ratio calculator and the percentage calculator convert between forms in a step. The full free online tools hub lists everything else.
Frequently Asked Questions
It sets out every possible combination of alleles that two parents can pass to an offspring, with one parent's gametes across the top and the other's down the side. Because each square is equally likely, counting squares gives the probability of each genotype and phenotype without further arithmetic.
Sixteen. Each doubly heterozygous parent produces four different gametes, and four multiplied by four gives a four-by-four grid. A monohybrid cross has four boxes and a trihybrid cross has sixty-four.
Because two independent 3:1 ratios multiply together. Nine of the sixteen squares are dominant at both genes, three are dominant at only the first, three at only the second, and one is recessive at both. It holds only when the two genes assort independently.
The genotype ratio counts the actual allele combinations, such as 1 AA : 2 Aa : 1 aa. The phenotype ratio counts what is visible, so under complete dominance AA and Aa look identical and the same cross gives 3 dominant : 1 recessive.
No. Eye colour is influenced by several genes and does not follow the single-gene, brown-dominant model taught in schools. This calculator is built for classic Mendelian teaching examples and should not be used to predict any real person's traits.
No, and it should not be. Real inheritance involves carrier status, variable penetrance, new mutations and family-specific variants that no grid captures. Questions about an actual family belong with a clinical geneticist or genetic counsellor.
It makes the heterozygote visibly intermediate, so the phenotype ratio matches the genotype ratio. A monohybrid cross of two heterozygotes gives 1:2:1 for both rather than the usual 3:1 phenotype split. This calculator assumes complete dominance.
No. It means each offspring independently has a three-in-four chance of showing the dominant phenotype. Small numbers of offspring often depart from the expected split, in the same way that four coin tosses often fail to give exactly two heads.
This calculator is an educational tool for biology coursework using classic Mendelian examples. It is not a genetic test, a diagnostic instrument, or a source of medical advice, and it must not be used to estimate any real person's traits or their risk of inheriting a condition. Anyone with a question about inheritance in their own family should speak to a qualified clinical geneticist or genetic counsellor.
