Wild horses first set foot in Europe roughly 2.6 million years ago, crossing from North America into Eurasia over land bridges that periodically connected the two continents. That migration introduced the genus Equus to Europe for the first time, but it was only one chapter in a much longer story. Horses originated in North America tens of millions of years earlier, disappeared from Europe entirely after the last ice age, and then returned as domesticated animals around 4,000 years ago from the steppes of western Eurasia. Untangling when horses “came to” Europe depends on which arrival you mean, and each one reshaped the continent in different ways.
Where Horses Actually Evolved
The horse family tree is rooted in North America, not Europe or Asia. The earliest known ancestors of modern horses were small, multi-toed browsers that appeared during the Eocene epoch, over 50 million years ago. These animals looked nothing like today’s horses. They were roughly the size of a dog, had padded feet with multiple toes, and ate soft leaves rather than grass. Over tens of millions of years, North American horse lineages gradually shifted toward larger body sizes, longer legs, and teeth better suited for tough vegetation. Early lineages like Hyracotherium through Mesohippus evolved their teeth relatively slowly, but during a major shift from browsing to grazing in the early Miocene, tooth crown height increased rapidly, reflecting a diet increasingly dominated by abrasive grasses.1Biological Journal of the Linnean Society. Fossil horses from “Eohippus” (Hyracotherium) to Equus. 2. Rates of dental evolution revisited
Throughout this long evolutionary arc, horse ancestors repeatedly crossed between North America and Eurasia via the Bering land bridge, which emerged and submerged as sea levels fluctuated. Some of these migrants gave rise to lineages that thrived briefly in the Old World before dying out. But the truly consequential crossing came when the genus Equus itself, the lineage that includes all modern horses, zebras, and donkeys, made the journey.
The First Equus in Europe
The earliest known appearance of Equus in Europe dates to about 2.6 million years ago, at the very start of the Pleistocene. Fossil sites across the continent tell a consistent story. At Roca-Neyra in central France, re-analysis of the equid fossils revealed the presence of Equus cf. E. livenzovensis alongside the last surviving hipparions, a more primitive three-toed horse group. The same species has been found at localities of similar age in Italy, Spain, and the Azov Sea region, establishing that the monodactyl (single-toed) horse spread across a wide swath of Europe within a geologically short window.2Journal of Paleontology. New insights on the Early Pleistocene equids from Roca-Neyra (France, central Europe): implications for the Hipparion LAD and the Equus FAD in Europe
The teeth of E. livenzovensis show features intermediate between the North American Pliocene species Equus simplicidens and the later European Equus stenonis, supporting the idea that this species was a stepping stone between the horse’s American homeland and its European descendants.2Journal of Paleontology. New insights on the Early Pleistocene equids from Roca-Neyra (France, central Europe): implications for the Hipparion LAD and the Equus FAD in Europe The arrival of Equus in Europe was part of a broader wave of animal movements known informally as the “Elephant-Equus event,” when large mammals poured into Eurasia from the Americas as conditions shifted. Hipparions had been present in Europe for millions of years before this, but they were progressively replaced by the more modern Equus lineage and eventually vanished.
Ice Age Horses Across Europe
Once established in Europe, wild horses became one of the continent’s most successful large mammals. Through the dramatic climate swings of the Pleistocene, they persisted across a range of habitats, from open steppe to scrubby woodland edges. Multiple species and subspecies occupied Europe simultaneously, and the genetic picture was remarkably diverse.
Ancient DNA work on wild horses from Switzerland shows that populations living during and directly after the Last Glacial Maximum (roughly 20,000 years ago) were genetically very distinct from one another, with large differences between groups. After the ice retreated, a highly diverse horse population expanded rapidly into newly available territory. This pattern was specific to Europe. In Asia, horse populations actually declined during the same period, making Europe’s post-glacial horse boom unusual.3PubMed Central. Ancient mtDNA diversity reveals specific population development of wild horses in Switzerland after the Last Glacial Maximum
Some of the most vivid evidence for Europe’s Pleistocene horses comes from cave art. The famous painted horses of Lascaux, Chauvet, and Pech Merle have long fascinated researchers, but a question lingered: were those spotted and dappled horses artistic fantasy, or did they actually look like that? Genetic analysis of 31 predomestic horse samples from Siberia, Eastern and Western Europe, and the Iberian Peninsula found that all the coat-color phenotypes visible in cave paintings existed in real horse populations. Eighteen were bay, seven were black, and six carried an allele for leopard complex spotting, the only spotted pattern identified in wild predomestic horses so far. Leopard spotting appeared in four Pleistocene samples and two Copper Age samples, confirming that the cave artists were painting what they saw rather than inventing patterns.4PubMed Central. Genotypes of predomestic horses match phenotypes painted in Paleolithic works of cave art
Europe was not limited to the ancestors of modern horses, either. The European wild ass (Equus hydruntinus) was another equid that persisted in Eurasia well into the Holocene, joining wild horses among the open-habitat megafauna that outlasted the Last Glacial Maximum in various refugia across the continent.
Extinction in the Americas, Survival in Eurasia
While European wild horses thrived through the ice ages, their relatives back in North America were heading toward oblivion. By the end of the Pleistocene, every horse species on the American continent had vanished. Recent work has clarified that this was not a drawn-out, staggered process. At least two distinct horse groups, the stout-legged caballine horses (closely related to modern horses) and the slender-limbed Haringtonhippus, lived side by side in both North America and Beringia through the late Pleistocene. New radiocarbon dates show that Haringtonhippus survived until about 14,400 years ago in eastern Beringia, meaning it likely overlapped with arriving human populations. Both horse groups then went extinct at roughly the same time across their entire North American range, suggesting that similar forces, whether climate change, human hunting, or both, drove them to extinction simultaneously.5PubMed Central. A new genus of horse from Pleistocene North America
In Eurasia, wild horses survived, though their ranges contracted as forests expanded during the warmer Holocene. By the time humans began domesticating horses, wild populations still dotted the steppes of central Eurasia, parts of the Iberian Peninsula, and scattered pockets elsewhere. The European wild horse would eventually disappear too, with the last truly wild populations vanishing in historical times, but its long Holocene survival set the stage for one of the most transformative human-animal partnerships in history.
Domestication and the Genetic Turnover
The question of where and when horses were first domesticated has been one of the most contentious debates in archaeology. For years, the Botai culture of northern Kazakhstan, dating to around 3500 BCE, was considered the leading candidate. Botai people kept horses in corrals, consumed mare’s milk, and seemingly relied on horses economically. But genomic analysis overturned the narrative. Compared to dozens of ancient and modern horse genomes, the Botai horses turned out to be the ancestors of Przewalski’s horses, not the ancestors of modern domestic horses. Today’s domesticates carry only about 2.7% Botai-related ancestry, meaning a massive genetic turnover replaced the Botai lineage almost entirely.6PubMed. Ancient genomes revisit the ancestry of domestic and Przewalski’s horses
A landmark 2021 study analyzing 273 ancient horse genomes pinpointed the lower Volga-Don region of the western Eurasian steppes as the homeland of the lineage that gave rise to modern domestic horses, often called DOM2. From about 2000 BCE onward, this lineage expanded rapidly across Eurasia, replacing almost all other local horse populations along the way. The expansion happened in lockstep with equestrian material culture, including the spoke-wheeled chariots associated with the Sintashta culture.7PubMed. The origins and spread of domestic horses from the Western Eurasian steppes
The timing of “true” domestication remains debated, though. Some researchers have argued that domestication should be equated with the appearance of certain favorable genetic mutations first evident in DOM2 horses around 2200-2100 BCE. Others push back against this genetic-centric definition, pointing out that horses from multiple genetic backgrounds, including lineages that predate DOM2, were managed, milked, and ridden well before that date. Evidence from archaeology, osteology, and ancient DNA suggests that people were exploiting horses in various ways long before the specific DOM2 lineage achieved its explosive spread.8PubMed Central. Horse genetics, archaeology, and the beginning of riding
Domestic Horses Arrive in Europe
If the first wild Equus reached Europe 2.6 million years ago, the arrival of domestic horses is a much more recent event, measurable in thousands rather than millions of years. The DOM2 lineage spread from the Pontic-Caspian steppes westward into Europe during the early to mid-second millennium BCE, carried by migrating peoples and traded between communities. By the time this lineage reached the western edges of Europe, it had already absorbed or displaced most local horse populations along its path.7PubMed. The origins and spread of domestic horses from the Western Eurasian steppes
Europe was not a genetic blank slate for horses when DOM2 arrived. The Iberian Peninsula, for example, harbored its own distinct horse lineage during the third and early second millennia BCE. Genomic analysis reveals that these native Iberian horses clustered separately from Przewalski’s horses, from the Botai lineage, and from the DOM2 lineage, representing a fourth horse lineage that existed during the early phase of domestication. Members of this group possessed their own distinctive mitochondrial DNA haplogroup and have been found at Spanish Chalcolithic and Bronze Age sites with archaeological contexts consistent with either wild or domestic status.9Cell. Ancient Genomes Revise the Evolutionary History of Domestic Horses What happened to this Iberian lineage as DOM2 swept across the continent remains an open question, but its genetic contribution to modern horses appears minimal.
In southern Scandinavia, ample archaeological evidence of close human-horse interaction appears by around 1600 BCE, including the famous Trundholm Sun Chariot from Denmark and chariot imagery from the Kivik grave in southern Sweden. Horses at this stage were primarily draught animals, pulling chariots. Mounted riding did not become widespread in Scandinavia until the transition from the Bronze Age to the Iron Age, around 700-500 BCE, when rock carvings of horseback riders and mounted warriors appear as part of a broader phenomenon spreading across Europe and Asia.10Trivent Publishing. Bridles and Bones: Early Cavalry in Southern Scandinavia
What Humans Selected for in the Horse
Domestication did not just move horses around the map. It reshaped them biologically. Genome-wide scans for positive selection in domestic horses have identified roughly 125 genes that appear to have been targeted during domestication. One cluster relates to muscular and limb development, joint structure, and the cardiovascular system, representing likely adaptations to the physical demands of riding, pulling, and carrying loads. A second cluster involves cognitive traits: social behavior, learning ability, fear response, and what researchers describe as agreeableness, traits that would have been crucial for taming and working with a large, powerful animal.11PubMed Central. Prehistoric genomes reveal the genetic foundation and cost of horse domestication
More recent breed-level analysis has confirmed selection signatures for coat color, body size, racing performance, gait patterns, and behavioral traits across European and Near Eastern horse breeds, along with less expected targets involving embryonic development, energy metabolism, and hair follicle formation.12PubMed Central. Genome-Wide Homozygosity Patterns and Evidence for Selection in a Set of European and Near Eastern Horse Breeds
Coat color tells a particularly interesting story about shifting human preferences over time. Spotted coat patterns, including sabino and tobiano, increased in frequency during the late Bronze and Iron Ages. Chestnut coloring also became more common during the Iron Age, showing signs of positive selection, while the trend for the ancestral bay and black phenotypes turned negative. These fashions in horse color were not static. The popularity of spotted patterns declined again during the Middle Ages, suggesting that human aesthetic preferences for horse appearance have swung back and forth over the millennia.13Scientific Reports. Spotted phenotypes in horses lost attractiveness in the Middle Ages
Przewalski’s Horses and the Question of “Truly Wild”
Przewalski’s horse, the stocky, dun-colored horse native to the Mongolian steppe, was long considered the last surviving wild horse species. Genomic evidence has complicated that picture considerably. The Botai-era horses of Kazakhstan, once thought to be early domestic ancestors of modern horses, instead turned out to be the progenitors of Przewalski’s horse. This means Przewalski’s horses are technically the feral descendants of horses that were herded by Botai people around 5,500 years ago, not a lineage that was never managed by humans.6PubMed. Ancient genomes revisit the ancestry of domestic and Przewalski’s horses
That said, the genetic divergence between Przewalski’s and domestic horses is real and ancient. Genomic analyses estimate that the two populations split roughly 45,000 years ago, though gene flow continued between them afterward.14PubMed Central. Evolutionary Genomics and Conservation of the Endangered Przewalski’s Horse An exome-based phylogeny places domestic and Przewalski’s horses as two distinct clades, both separate from late Pleistocene horses. Demographic modeling suggests their population trajectories ran in parallel until relatively recently, with an early divergence for the Pleistocene horse population at around 150,000 years ago.15Current Biology. Evolutionary Genomics and Conservation of the Endangered Przewalski’s Horse Sex chromosomal and autosomal DNA analyses further support a close relationship between the two groups, with Przewalski’s horses not forming a clearly separate clade from domestics on X chromosomal or autosomal trees.16Molecular Biology and Evolution. Horse Domestication and Conservation Genetics of Przewalski’s Horse Inferred from Sex Chromosomal and Autosomal Sequences
The upshot is philosophically awkward for conservation: if “wild” means “never domesticated,” then there may be no truly wild horses alive today. But if “wild” means “free-living, self-sustaining, and genetically distinct,” Przewalski’s horse still qualifies. The label matters less than the animal’s ecological role and genetic uniqueness.
Reading the Bones and Teeth
Fossil bones and teeth are more than species identifiers. They record how horses lived and how humans used them. Bioarchaeological study of Iron Age horses from Thrace (modern Bulgaria) found dental wear patterns consistent with bit use, while vertebral and phalangeal lesions indicated prolonged and repetitive mechanical loading, the kind of damage you would expect from sustained draught work and load-bearing over a horse’s lifetime.17PubMed. Early Iron Age horse exploitation in Thrace (Bulgaria) inferred from dental attrition and vertebral pathology Findings like these help fill in the gap between “horses were present” and “horses were being worked hard,” offering tangible evidence of how deeply embedded equine labor had become in European Iron Age economies.
Biomolecular evidence adds another dimension. Paleo-proteomic analysis of dental calculus from Iron Age Scythian individuals detected horse milk protein (beta-lactoglobulin 1) in a single person at the Bilsk site. While Scythian reliance on ruminant dairy was already well established, this provided the first direct biomolecular evidence that Scythians also consumed horse-derived products, confirming literary accounts of their horse-centered pastoral culture.18PLoS One. Paleo-proteomic analysis of Iron Age dental calculus provides direct evidence of Scythian reliance on ruminant dairy
Rewilding and the Return of Free-Ranging Horses
In a twist that brings the story full circle, horses are now being reintroduced into European landscapes as part of rewilding programs. Across the continent, cattle and horse breeds are being placed on reserves as functional substitutes for the extinct aurochs and wild horse (tarpan) that once shaped European grasslands and woodlands through their grazing.19Restoration Ecology. Rewilding Europe’s large grazer community: how functionally diverse are the diets of European bison, cattle, and horses? The idea is that large herbivores are needed to maintain the open and semi-open habitats that many European species depend on, and that domestic breeds with relatively wild-type behavior can fill the ecological niche left vacant since the Holocene extinctions.
Central to this effort is a process sometimes called “de-domestication,” selecting horse breeds that retain hardiness, strong social structures, and self-sufficient foraging behavior, and managing them with minimal human intervention.20PubMed Central. De-extinction beyond species: Restoring ecosystem functionality through large herbivore rewilding Breeds commonly used include Koniks, Exmoors, and various primitive-type ponies that tolerate harsh conditions. The integration of horses into European rewilding has gained widespread popularity in recent decades, driven by evidence of their potential for regulating vegetation and restoring natural ecosystem dynamics.21PubMed Central. Rewilded horses in European nature conservation – a genetics, ethics, and welfare perspective
The practice raises thorny questions about genetics, ethics, and animal welfare. These rewilded horses are not genetically wild. They carry the signatures of thousands of years of selective breeding, including the cognitive and behavioral traits that made their ancestors manageable in the first place. Whether a de-domesticated horse can truly replicate the ecological behavior of a Pleistocene wild horse, or whether it simply approximates it well enough to be useful, is a question rewilding ecologists continue to wrestle with. The horses themselves, meanwhile, seem unbothered by the debate. On reserves from the Netherlands to Romania, free-ranging herds graze, form social bands, and reshape the landscape much as their ancestors did hundreds of thousands of years before any human ever thought to throw a leg over one.