Life › Human evolution › Hominins: the human branch begins
Walking on two legs
by c. 6 million years ago; details debated· Evolutionary innovation· Human ancestors & us
Strong evidenceGood evidence; dates or details still being refined.
Habitual walking on two legs, the trait that first set the human line apart from other apes.
See it on the Life map →What happened
Other apes mostly climb or knuckle-walk; hominins walk upright. The change reshaped the skeleton: an S-curved spine, a short bowl-shaped pelvis, thighbones angled inward toward the knees, and arched feet with the big toe in line. Early hominins such as Ardi, 4.4 million years ago, still had grasping big toes for climbing. Largely human-like striding shows in the Laetoli footprints, 3.66 million years old.
Why it matters
It freed the hands for carrying and later tool-making, and made long-distance walking cheap.
The evidence (fossils, genes, rocks)
Fossil hips, thighbones and feet (Orrorin, Ardi, Lucy), the Laetoli footprints, and lab studies (2007) showing human walking uses about 75% less energy than chimpanzee walking.
Still debated
Why it evolved is debated: open grassland, energy savings, carrying food, staying cool, or reaching fruit while standing in trees. Fossil apes such as Danuvius (2019) complicate the picture.
Inside this (1)
Connections (7)
▸Grasslands spread may have favored Walking on two legsDebated
The connection
The classic 'savanna hypothesis' holds that as African forests broke up into open grassland, ancestors who walked upright could see farther, travel efficiently between patches of trees and stay cooler under the sun. On this view, upright walking began as an adaptation to open country.
When and how
East and Central Africa, about 7–4 million years ago; the idea dates to the early 1900s and was revived by Yves Coppens' 'East Side Story' (1994).
How sure are we?
Uncertain. Fossil soils, plants and animals suggest early hominins such as Ardi, 4.4 million years ago, lived in wooded settings, not open savanna, though some researchers dispute this.
Debates and open questions
Many researchers now think bipedalism began in mixed woodland, with open grasslands mattering more later, for Homo. How open early hominin habitats were is still argued.
▸Sahelanthropus tchadensis may show early signs of Walking on two legsDebated
The connection
Two lines of evidence suggest Sahelanthropus may have walked upright at times. The opening at the skull base where the spinal cord exits appears positioned forward, as in upright walkers, and a thighbone from the same site may show bipedal features. It probably also climbed trees using strong arms.
When and how
Toros-Menalla, Chad; skull described in 2002, thighbone and forearm bones described in 2022 by Guillaume Daver, Franck Guy and colleagues.
How sure are we?
Skull-base angle and thighbone shape compared with apes and humans; the thighbone is assumed to belong to the same species, though it was found apart from the skull.
Debates and open questions
A 2020 analysis of the same thighbone concluded it was not from an upright walker; later studies found hip-muscle features typical of bipeds. Opinion remains split.
▸Hominins: the human branch begins are defined by Walking on two legsStrong evidence
The connection
Scientists usually identify a fossil as a hominin by signs of habitual upright walking, along with smaller canine teeth. Because skulls, hips and legs are rarely found together, each early candidate is judged by whatever bones survive: Orrorin's thighbone, Sahelanthropus's skull base, and Ardipithecus's pelvis and feet.
When and how
Africa, from about 7–6 million years ago; key early evidence comes from Orrorin (Kenya, described 2001), Ardi (Ethiopia, described 2009) and later Lucy.
How sure are we?
Fossil hips, thighbones and feet show the anatomy of upright walking at least by 4.4 million years ago; earlier candidates rest on fewer, disputed bones.
Debates and open questions
Some fossil apes, such as Danuvius from Germany, about 11.6 million years old (described 2019), may have stood upright too, blurring bipedalism as a strictly hominin trait.
▸Ardipithecus ramidus ('Ardi') mixes climbing with early Walking on two legsStrong evidence
The connection
Ardi's pelvis and feet suggest a creature that walked upright on the ground but still climbed easily. Her big toe was splayed for grasping branches, yet her upper pelvis was reshaped for upright walking. To her finders, this mix shows two-legged walking began in a tree-climbing ape in woodland, not in a savanna runner.
When and how
Aramis, in Ethiopia's Middle Awash valley, about 4.4 million years ago; skeleton excavated from 1994 by Tim White's team and described in 2009.
How sure are we?
The partial skeleton includes pelvis, hands and feet. But the pelvis was crushed and rebuilt digitally from scans, so readings of how she walked rest on reconstruction.
Debates and open questions
Some anatomists doubt the pelvis reconstruction and question how upright Ardi really walked; others doubt that bipedalism began in woodland, as her finders argue.
▸Lucy (Australopithecus afarensis) shows a skeleton built for Walking on two legsWell established
The connection
Lucy's pelvis is short and wide, and her thighbones angle inward toward the knees, as ours do, placing the feet under the body's center. These are signs of habitual upright walking, yet her long, curved fingers and strong arms suggest she still climbed trees.
When and how
Hadar, Ethiopia: a separate knee joint found there in 1973 first showed the angled thighbone, and Lucy's skeleton, found in 1974, confirmed it.
The evidence
Comparisons with human and ape skeletons, and CT scans of her bones' internal structure, which reflects the loads they carried in life, support upright walking.
Debates and open questions
How efficiently she walked compared with us, and whether her species' climbing traits were still useful or simply leftovers from tree-dwelling ancestors.
▸Laetoli footprints record Walking on two legsWell established
The connection
Footprints capture behavior, not just anatomy. The Laetoli trails show a heel striking first, a raised arch, and a big toe in line with the other toes: a walking style close to ours, already in place about 3.66 million years ago.
When and how
Laetoli, northern Tanzania, about 3.66 million years ago; the main trail was found in 1978 by Mary Leakey's team, with more prints added in 2015.
The evidence
Experiments comparing the prints with those of people and chimpanzees walking in wet sand point to an upright, humanlike stride rather than a bent-knee ape gait.
Debates and open questions
Studies differ on how humanlike the stride was; some find a slightly bent knee or a less efficient walk than ours.
▸Walking on two legs freed hands for First stone toolsStrong evidence
The connection
Walking upright freed the hands from getting around. Over millions of years, hominin hands evolved shorter fingers and longer thumbs, allowing a precision grip for holding a stone and striking it at the right angle. Tools came some 3 million years after bipedalism, so walking alone did not cause them.
When and how
Africa: bipedalism by about 6 million years ago; first stone tools about 3.3–2.6 million years ago.
How sure are we?
Fossil hands show precision-grip features increasing over time; experiments measure the grip and forces needed to strike sharp flakes.
Debates and open questions
Darwin proposed this link in 1871. Why there was such a long delay, and whether earlier tools of wood or unmodified stone left no trace, is unknown.