In short: enter the genetic results or the visible coats of the stallion and the mare, and the simulator immediately works out the probability of every possible coat for the foal. It is as useful to an experienced breeder holding DNA results as to an owner who only knows what colour their horses are.
Genotype mode is for those who have had a DNA test run on their breeding animals. The laboratory then supplies precise results: EE, Ee, Aa, Crn, and so on. These codes describe exactly which alleles the horse carries, meaning the two versions it holds of each gene. It is the most reliable way to predict a foal's coat, because it reveals silent carriers, those horses that pass on a gene without showing it. A chestnut carrying Cream, for instance, can produce a palomino with nothing about its own appearance giving it away.
Declared coat mode suits you when all you know is the coat you can see. Describe it gene by gene: base colour (bay, black or chestnut), dilutions, graying and patterns. Every combination the engine can compute is declared that way, with no DNA test, a dun perlino and a silver buckskin included.
The simulator works out the compatible genotypes and computes the possible combinations. Coats such as « palomino » or « buckskin » are therefore composed results rather than entries to pick from: the name is built out of the genes you tick.
Without a DNA test there is no way to know which genotype the horse actually carries, so the simulator treats the compatible genotypes as equally likely. That is a convention of calculation, not a frequency observed in a breed. In the « The genotypes behind this coat » panel the rows are therefore ranked from the most likely to the least likely, while the coat probabilities are still shown, rounded and prefixed with « ≈ ».
Uncertainty is accordingly greater than in genotype mode, since the silent carriers mentioned above stay invisible.
The special case of the gray horse. Gray has a row of its own, to be ticked, rather than an entry in a list of coats. The official IFCE nomenclature does list gray among the base coats, beside white, cream and chocolate, but it describes what an identifier sees at a given moment. This simulator has to predict what the foal will inherit, and graying progressively erases the birth coat without replacing it: a gray horse was bay, chestnut or black before it whitened, and that is the colour it passes on. This is why every gray result is written in two parts, « Gray » and, just below it, « born bay ».
So tick the Graying row, and give the birth coat on the « Base coat » row if you know it. If you do not, which is the usual case, answer « I do not know »: the calculation still runs and the simulator explores every possible base. Two useful details. The copy number of the gray gene can now be declared, one or two, and it changes what the parent passes on: one copy gives a gray foal half of the time, two copies give one every time. And gray stacks freely with a dilution, so a gray born palomino is described here without losing anything.
One feature the coat composer does not carry. Traits that do not change the background colour are declared nowhere in this tool, and that applies in particular to a flaxen mane and tail, meaning a mane and tail lighter than the body, a feature that is nevertheless written in full on French registration papers. The reason is easy to state: the gene responsible has not been identified yet, it is most likely spread across several genes at once, and no laboratory offers a test for it. The simulator cannot pass it on, so it prefers to promise nothing. The same goes for sooty, pangaré and rabicano. Declare the background colour alone: a “flaxen chestnut” is declared here as a chestnut, with nothing else ticked.
The two modes are chosen separately for each parent. If you have tested your stallion but not your mare, set the first panel to genotype and the second to phenotype: the simulator combines the two without difficulty and reports in its results which assumptions it had to make. Be aware, though, that a single parent described by its coat is enough to switch the whole calculation to approximate mode. Coat percentages are then shown with the « ≈ » sign and the per-genotype detail is ranked rather than numbered, including for the parent you did have tested, since every final probability depends on both parents at once.
Coat genetics rests on a clear hierarchy: two base genes set the fundamental colour, then dilution genes and white markings modify it. The simulator takes 21 loci into account, which covers the coat genes found on the panels laboratories commonly offer. The ten main ones are detailed below; the other eleven, which cover rare white markings, appaloosa patterns and coat texture, are available in the “Advanced genes” section of each input panel.
| Gene | What it controls | Coats produced, as the simulator names them | Associated health note |
|---|---|---|---|
 Extension (E / e) | What it controlsProduction of eumelanin, the black pigment. e/e gives a red base and nothing else. | Coats produced, as the simulator names themChestnut (e/e), Bay or Black (E/_) | Associated health notenone |
 Agouti (A / a) | What it controlsHow black is distributed over the body. Acts only when E is present. | Coats produced, as the simulator names themBay (A/_ with E/_), Black (a/a with E/_) | Associated health notenone |
 Cream (N / Cr / prl) | What it controlsDilutes the base coat, with a stronger effect in a double dose. The same locus carries the pearl allele. | Coats produced, as the simulator names themPalomino, Buckskin, Cremello, Perlino, Smoky Cream, Pearl Chestnut | Associated health notenone |
 Dun (D / nd1 / nd2) | What it controlsDilutes the body and adds primitive markings: a dorsal stripe and leg barring. The two non-dun forms give the same coat, but nd1 still permits the primitive markings where nd2 removes them, which the simulator flags. | Coats produced, as the simulator names themBay Dun, Red Dun, Grullo | Associated health notenone |
 Silver (N / Z) | What it controlsWashes out the black pigment of the body, mane and tail. No visible effect on a chestnut. | Coats produced, as the simulator names themSilver Bay, Silver Black | Associated health noteEye anomaly (MCOA). Severe form in Z/Z, an isolated cyst in Z/N. |
 Champagne (N / Ch) | What it controlsDominant dilution of both pigments, with mottled skin and light eyes at birth. | Coats produced, as the simulator names themGold Champagne, Amber Champagne, Classic Champagne | Associated health notenone |
 Gray (N / G2 / G3) | What it controlsProgressive graying with age. Careful, the two forms are not equivalent: G3 leads to white, whereas a G2 horse stays dappled and never turns white. | Coats produced, as the simulator names themGray, on every base. The result keeps the birth coat as a subtitle, « born bay » for instance. | Associated health noteHigh melanoma risk in G3/G3. |
 Roan (N / Rn) | What it controlsWhite hairs mixed through the body, stable throughout life. | Coats produced, as the simulator names themBay Roan, Red Roan (a roan on a chestnut base), Blue Roan. Roan becomes a mere suffix as soon as a dilution takes the head of the name, as in « Buckskin dun roan ». | Associated health notenone |
 Tobiano (N / TO) | What it controlsDominant pinto marking: large rounded white patches, often crossing the back. | Coats produced, as the simulator names themTobiano, on every base | Associated health notenone |
 Frame Overo (N / O) | What it controlsPinto marking: ragged patches on the flanks, with the head often white. | Coats produced, as the simulator names themFrame Overo, in the heterozygous state N/O only | Associated health noteLethal in the homozygous state O/O, see below. |
Two genes in the “Advanced genes” section also carry a health note: the leopard complex, meaning the gene that produces appaloosa patterns (congenital stationary night blindness in LP/LP, plus a risk of uveitis), and the Curly locus KRT25 (hypotrichosis, meaning very reduced hair growth). These notes are flagged risks, not diagnoses: only a vet can say anything about a particular horse.
If your horse is a solid appaloosa. An appaloosa-registered horse whose coat carries no contrasting pattern is called “solid” in registry language, and two opposite situations have to be told apart. If it shows mottled skin around the muzzle, eyes and genitals, visible white sclera around the iris, or vertically striped hooves on a leg with no white marking, then it does carry the gene and you must tick the “Appaloosa pattern” box. If it shows neither a pattern nor any of those three signs, which registries record as “non-characteristic”, it does not carry it and the box must stay empty. Ticking it wrongly adds a uveitis risk to every one of your matings; failing to tick it when you should removes the night-blindness warning altogether.
A foal that is homozygous O/O, born to two Frame Overo carriers, is born white and has Overo Lethal White Syndrome (OLWS). The intestinal aganglionosis that comes with it, meaning the absence of the nerve cells that drive the gut, is not survivable. If both of your breeding animals carry Frame Overo, the risk is 25 % for every foal. A DNA test beforehand is essential before any mating that involves this gene.
That figure of 25 % assumes both parents have been declared carriers. You need to understand how the simulator treats a gene you have not filled in: it assumes it is absent. Put another way, the absence of a warning does not mean the absence of a risk, it means the question was never asked. If you do not know the Frame Overo status of one of the two horses, choose the “Not tested” value rather than leaving the field on its default: the simulator will then show a risk range instead of a misleading zero. This is the one place in this tool where a default assumption can cost a foal.
The percentages shown are statistical probabilities, computed gene by gene with Punnett squares and then combined. Concretely, if the simulator says 60 % bay, that means about six out of ten foals from this mating would be bay. It does not mean the next foal will be bay. When one of the two parents was described by its coat rather than tested, those percentages also rest on the equal-likelihood assumption described above, and the “≈” sign in front of them is there precisely as a reminder.
Every birth is an independent draw. The same two parents can perfectly well produce three chestnuts in a row before giving a bay, even when bay is the most likely outcome. Horse coat genetics follows Mendel's laws, which are laws of probability and not of certainty.
What the simulator cannot do: confirm the actual genotype of an individual horse. Only a DNA test run by a veterinary laboratory gives that certainty. The simulator is a planning and exploration tool. It helps you predict a foal's coat and spot the interesting or risky combinations before the mating even happens.
The phenotype is what you see, the horse's visible coat. The genotype is the actual combination of alleles the animal carries, which only a DNA test can reveal. A horse can have a chestnut phenotype and a Cream-carrying genotype: same visible coat, very different mating results.
Yes, for statistical probabilities: the calculations rest on Mendelian genetics and Punnett squares, which are the reference method. No, for predicting one particular foal's coat with certainty, because every birth remains a random draw. In declared coat mode, reliability is further limited by silent carriers that have not been detected and by the equal-likelihood assumption across the genotypes compatible with the declared coat. That is why percentages are rounded there and the per-genotype detail is presented as a ranking.
A bay carries at least one E allele (Extension) and one A allele (Agouti). So it can be E/E or E/e, and A/A or A/a. Without a DNA test you cannot know which. Do not try to guess and do not run two calculations to compare them: choose the “Not tested” value each selector offers. The simulator then handles both possibilities at once and gives you a range of probabilities, which is more honest than a single, falsely precise figure.
No, never, and this is one of the few certainties in coat genetics. A chestnut is e/e, so it has no E allele to pass on, and without E the black pigment is not produced. Two chestnut parents necessarily produce a chestnut foal, possibly diluted to palomino or cremello if a dilution gene is involved, but never bay and never black.
It is a horse that carries an allele without showing it. The most common example is a chestnut carrying Cream, of genotype Crn: it has a perfectly ordinary coat, but it can produce a palomino or a buckskin if the other parent also carries the gene or expresses the dilution. These carriers are invisible to the eye and only a DNA test reveals them, which is exactly why genotype mode is worth using.
Through a DNA test run by a specialist veterinary laboratory. Many laboratories offer complete coat genetics panels, and your studbook can usually tell you which one it works with. Équipédia, the online resource of the IFCE, the French horse and riding institute, is cited at the foot of this page and is a solid reference for understanding which genes can be tested and choosing the right panel for your breed and your breeding goals, although it is written in French.
Gene by gene, and with no DNA typing. A dun perlino, that is a horse with a bay base coat carrying two copies of the cream gene (which gives it a very pale coat and blue eyes) and the dun gene (which adds a dark stripe along the back), is declared in three moves: bay base coat, cream gene in two copies, dun gene. The simulator then composes the name itself and shows it at the top of the input block, « Perlino dun », and computes the cross from that genotype. No combination needs to exist in a list beforehand: the name comes out of the stack of genes you tick, which holds just as well for a silver buckskin, a pearl red dun or a gray born palomino.
The odds on a foal's coat are only one criterion among many when choosing a sire, but they are often the one that starts the thinking. On Equ'un, the page of a stallion whose genetic profile has been filled in opens the simulator pre-loaded with his data: all that is left is to enter your mare's.
That is a real saving of time when comparing several stallions and narrowing down your choice before contacting a stud.
Search for a stallion on Equ'un
Emma Robert, co-founder of Equ'un and subject-matter lead. A rider for more than twenty years, she is the one who settles the questions of substance, from how a breeding season actually unfolds to the wording used.
LinkedIn profile
Maxence Fortin, co-founder and the chief technology officer of Equ'un. A software engineer, he co-created Equ'un to modernise how mare owners find a stallion and to simplify the paperwork that comes with it. His conviction: technology should serve the field, never the other way round.
LinkedIn profileLoading genetics catalogue…
