Horses were first domesticated in the western Eurasian steppe, in the region around the lower Volga and Don rivers of what is now southern Russia and western Kazakhstan, roughly 5,000 to 6,000 years ago. That origin story, however, took decades of debate and multiple DNA revolutions to pin down. An earlier candidate site in Central Asia turned out to be a dead end, and the horses that eventually became every modern breed on Earth replaced all other local horse populations in a remarkably fast sweep across two continents.
The Lower Volga-Don Region
Genetic analysis of dozens of ancient horse remains has traced a single ancestral lineage, referred to in research as DOM2, back to the lower Volga-Don region. This lineage first appears in archaeological contexts associated with the late Yamnaya culture, a Bronze Age pastoralist society that spread across the Pontic-Caspian steppe. By around 2200 BCE, DOM2 horses had become geographically widespread, and within just a few centuries they dominated horse populations from Western Europe to East Asia.
What makes the DOM2 story so striking is its completeness. This wasn’t a gradual blending of many local horse populations. It was a near-total replacement. As domesticated herds moved outward from the steppe, they were repeatedly restocked with local wild mares, which explains the wide diversity in mitochondrial DNA (inherited from mothers) alongside a narrow paternal lineage. But the core genome of every modern domestic horse traces back to that single steppe origin.
Why Botai Doesn’t Count
For years, the strongest archaeological case for early horse domestication came from the Botai culture in northern Kazakhstan, dating to around 3500 BCE. Botai settlements had huge quantities of horse bones, corrals, and ceramic vessels containing traces of mare’s milk fat. That milk residue, identified through carbon and hydrogen isotope analysis of fatty acids, pointed to a developed domestic economy where people were not just hunting horses but keeping and milking them.
Then came a surprise. When researchers sequenced 20 ancient genomes from Botai horses and compared them to modern breeds, the Botai animals turned out to be ancestors not of today’s domestic horses but of Przewalski’s horses, the stocky, wild-looking animals that survive in Mongolia. Modern domestic horses carry only about 2.7% Botai-related ancestry. Przewalski’s horses, long assumed to be the last truly wild horse species, are actually feral descendants of the Botai herds that were abandoned or escaped.
This finding reshuffled the entire timeline. The Botai people did domesticate horses, or at least managed them closely enough to milk them. But their horses were a different lineage that ultimately failed to become the foundation of modern horse culture. The massive genomic turnover that produced today’s domestic horses happened later, coinciding with large-scale human population expansions during the Early Bronze Age, centered further west in the Volga-Don steppe.
What Made DOM2 Horses Different
Researchers have identified specific genetic changes that were selected for during the earliest stages of domestication. One gene involved in modulating behavior shows a signature of positive selection beginning around 5,000 years ago, suggesting that tameness was one of the first traits early herders bred for. A calmer horse is easier to handle, breed, and keep in close quarters with people.
Shortly after, around 4,750 years ago, intensive selection began on a gene linked to body conformation, spinal anatomy, and muscular coordination. In mice, variants of this same gene affect motor strength and spinal structure. In horses, the selected variants reached high frequency by about 4,150 years ago. The implication is clear: early breeders were selecting for animals with stronger backs, better suited to carrying weight or pulling loads. These two traits together, docility and a robust spine, would have given DOM2 horses an enormous practical advantage over wilder, less physically suited populations.
From Herding to Chariots
The earliest use of domesticated horses was probably for meat and milk rather than riding or transport. The Botai evidence, despite its genetic dead end, shows that horse-based pastoral economies existed by the mid-fourth millennium BCE. In the Volga-Don region, early DOM2 horses likely served similar purposes before being ridden.
Direct evidence of bit use in equids dates to the early third millennium BCE. Archaeologists identify bit wear by examining the lower second premolar, which sits just behind the gap where a bit rests on the tongue and gums. A bit rubbing against this tooth creates distinctive enamel erosion that doesn’t appear in wild horses or domestic horses that were never bitted. The earliest well-dated example of this wear comes from a donkey skeleton at a site in Israel, dated to around 2800 to 2600 BCE, showing that equids were being controlled with bits by that period.
The real acceleration came with the chariot. The first confirmed horse-drawn chariots, light two-wheeled vehicles with spoked wheels, appear at the Sintashta-Petrovka archaeological site on the southern Urals, dating to roughly 2100 to 1700 BCE. These chariots increased travel speed across open steppe from about two miles per hour (walking with an ox cart) to ten. By around 1800 BCE, horse-drawn chariots had reached the Near East, as depicted on Syrian cylinder seals, and quickly became a decisive military technology.
What Happened to Other Wild Horses
The wild ancestor of modern domestic horses is extinct. As DOM2 horses spread across Eurasia, they replaced local wild populations through a combination of interbreeding and competition for grazing territory. The only surviving relative of those wild populations is Przewalski’s horse, which itself turns out to be feral rather than truly wild. The last Przewalski’s horse seen in nature was spotted in Mongolia’s Dzungarian Gobi desert in 1969. The species was saved through captive breeding programs and has since been reintroduced to reserves in Mongolia and China, but its genome tells the story of the Botai herds, not of the wild horses that once roamed the western steppe.
Why the Iberian Theory Fell Short
Some researchers once proposed that wild horses on the Iberian Peninsula were independently domesticated, based on the high frequency of a particular mitochondrial DNA cluster in modern Iberian breeds like the Lusitano. To test this, scientists sequenced mitochondrial DNA from 22 ancient Iberian horse remains spanning the Neolithic, Bronze Age, and Middle Ages. The Neolithic and Bronze Age horses grouped into different genetic clusters than the one overrepresented in modern Iberian breeds. Only a medieval Iberian sample matched the modern pattern, suggesting that the distinctive genetic signature of today’s Iberian horses arrived with already-domesticated stock rather than originating from local wild populations. The same story played out with other proposed independent domestication centers: when ancient DNA was tested, the local wild horses turned out to be genetic dead ends, replaced by the expanding DOM2 lineage from the western steppe.