Last updated: August 11, 2026
Every foal’s color is decided the moment it’s conceived, not when it’s born. What you see when a mare foals out is just the visible result of a genetic combination that was locked in nine months earlier. Understanding how that combination works is different from learning to identify a color at a glance – it’s the math behind why two bay parents can produce a chestnut foal, and why a breeder can sometimes predict color odds before the mare is even in foal.
Horse coat color starts with two genes working together – Extension (E/e) decides whether a horse can produce black pigment at all, and Agouti (A/a) decides where that black pigment shows up on the body. Together they produce the three base colors: black, bay, and chestnut. Dominant genes need only one copy to show; recessive genes need two. Everything else – dilutions, patterns, grays – is additional genes modifying that base.
I’ve bred and raised horses in Louisiana for over 30 years, and the truth is you don’t need a genetics degree to use this – you need to understand four or five real concepts well enough to make sense of a pedigree or predict a cross. Here’s the part that actually matters for an owner or breeder, without padding it into a biology course.
Table of Contents
How Inheritance Actually Works
DNA carries the instructions for every trait a horse has, coat color included. Sections of DNA called genes code for specific traits, and different versions of the same gene are called alleles. A horse gets one copy of most genes from its sire and one from its dam.
Not all alleles carry equal weight. A dominant allele only needs one copy to show up in the coat – if a horse has it from either parent, you’ll see its effect. A recessive allele needs two copies, one from each parent, or it stays hidden. That single distinction explains most of the surprises breeders run into: a horse can carry a color it doesn’t display, and pass it to a foal without ever showing it themselves.
The Two Genes Behind Black, Bay, and Chestnut
Two genes do almost all the work of deciding a horse’s base color, before anything else modifies it.
Extension (E/e): This gene decides whether black pigment can be produced at all. The dominant E allele allows it; the recessive e allele blocks it, forcing red pigment instead. A horse needs two copies of e (e/e) to be chestnut, since chestnut is what happens when black pigment simply isn’t available.
Agouti (A/a): This gene only matters if a horse can produce black pigment in the first place. The dominant A allele restricts that black to the points – mane, tail, and lower legs – which produces bay. The recessive a/a genotype lets black cover the whole body, producing true black.
| Extension | Agouti | Result |
|---|---|---|
| E/_ (at least one E) | A/_ (at least one A) | Bay – black restricted to the points |
| E/_ (at least one E) | a/a | Black – black covers the full body |
| e/e | Doesn’t matter | Chestnut – no black pigment produced |
Everything past this point – palomino, buckskin, dun, gray, roan, pinto, and Appaloosa patterns – is additional genes acting on one of these three foundations. For the full identification-focused breakdown of every color and pattern, see my complete horse coat colors guide.

Working Out a Real Cross
A Punnett square is just a grid that shows every possible combination of alleles two parents can pass to a foal. It’s the tool that turns “what color might this cross produce” from a guess into actual odds.
Take a bay mare who is heterozygous for Extension (Ee) bred to a black stallion who is homozygous (EE). Every foal gets an E from the stallion no matter what. From the mare, a foal has a 50% chance of getting E and a 50% chance of getting e. Since E is dominant, every possible foal in this cross ends up E/_ – meaning every foal can produce black pigment, and the actual bay-or-black outcome then comes down to what each parent carries at the Agouti gene.
| E (from stallion) | E (from stallion) | |
|---|---|---|
| E (from mare) | EE | EE |
| e (from mare) | Ee | Ee |
Miles’s Take: This is the part that actually pays off at the barn. I’ve had owners assume a cross is a “safe bet” for a certain color because both parents show it, without realizing one of them is heterozygous and carrying something else underneath. A quick Punnett square, or a DNA panel if the stakes are real, tells you the actual odds instead of a hopeful guess. It’s five minutes of work that can save a genuinely disappointing surprise nine months later.
Dilution Genes: Lightening the Base
Dilution genes act on top of the three base colors, lightening them into an entirely different family of colors. The three most common:
Cream: One copy turns chestnut into palomino, bay into buckskin, and black into smoky black. Two copies produce the much lighter “double dilutes” – cremello, perlino, and smoky cream.
Dun: Lightens the body while leaving primitive markings behind – a dorsal stripe, leg barring, and sometimes shoulder shading. Dun on black produces grullo; dun on bay produces bay dun; dun on chestnut produces red dun.
Champagne: Dilutes both red and black pigment, producing gold, amber, and classic champagne shades, usually paired with pinkish, freckled skin and light eyes.
For genuinely rare colors like silver dapple or champagne, a genetic panel is worth the cost before you plan a breeding around producing one – visual identification alone isn’t reliable enough to base a real breeding decision on.
Why Genetics Matters Beyond Curiosity
Understanding inheritance isn’t just a fun fact for the barn aisle. A few coat-color genes carry real health considerations worth knowing before you breed – frame overo and lethal white foal syndrome, and the leopard complex gene’s link to night blindness, among others. Genetics can also settle registration disputes when a color call is unclear, and it turns breeding decisions from hopeful guesswork into something closer to actual planning. For the specific health and breeding notes tied to individual colors and patterns, see my coat colors guide, and for the deeper genetic testing methods breeders actually use, see my advanced color genetics guide.
FAQs About Horse Color Genetics
What two genes determine a horse’s base color?
Extension (E/e) determines whether a horse can produce black pigment at all. Agouti (A/a) determines where that black pigment is placed on the body. Together, they produce the three base colors: bay, black, and chestnut.
Can two bay horses produce a chestnut foal?
Yes, if both parents are heterozygous for Extension (Ee). Each parent has a 50% chance of passing the recessive e allele, so two Ee parents have roughly a 25% chance of producing an e/e foal, which would be chestnut regardless of the parents’ own bay coloring.
What is the difference between a dominant and a recessive gene?
A dominant gene only needs one copy to show its effect – a horse with even one copy will display the trait. A recessive gene needs two copies, one from each parent, or its effect stays hidden. A horse can carry a hidden recessive allele and pass it to a foal without ever showing it themselves.
How is a Punnett square used in horse breeding?
A Punnett square maps out every possible combination of alleles two parents can pass to a foal, turning color prediction from a guess into actual odds. It’s most useful when at least one parent is heterozygous for a relevant gene, since the outcome isn’t automatically guaranteed either way.
When should I get a DNA color panel instead of guessing?
Get a panel before breeding for a specific rare color, when a registration color call is disputed, or when a color or pattern gene has a known health implication, like frame overo or the leopard complex gene. Visual identification is a reasonable starting point for casual purposes, but it isn’t reliable enough to base a real breeding or purchase decision on.
The math behind coat color isn’t complicated once you know the two genes that matter most – the rest of equine color genetics is really just variations built on that same foundation.

About Miles Henry
Racehorse Owner & Author | 30+ Years in Thoroughbred Racing
Miles Henry (legal name: William Bradley) is a professional horseman based in Folsom, Louisiana. He holds Louisiana Racing License #67012 and has spent over three decades managing Thoroughbreds at premier tracks including Fair Grounds, Delta Downs, and Evangeline Downs.
Expertise & Hands-On Experience: Beyond the track, Miles has decades of experience in specialized equine care, covering everything from hoof health and nutrition to training protocols for Quarter Horses, Friesians, and Paints. Every guide on Horse Racing Sense is rooted in this “boots-on-the-ground” perspective.
30 of their last 90 starts
Equibase Profile.
Connect with Miles:

