Yes. As far as the fossils show, our ancestors walked on two legs long before their brains got big. Footprints left in volcanic ash at Laetoli, Tanzania, about 3.66 million years ago are widely accepted as the oldest clear evidence of fully two-legged walking (McNutt et al. 2021). One of the best-known species from that time, Australopithecus afarensis, had an average brain volume of about 445 milliliters. A chimpanzee averages about 369, and a modern human brain is roughly three times a chimpanzee’s (Gunz et al. 2020). Clearly bigger brains show up later, in some early members of our own genus, Homo: one reconstructed skull from about 1.8 million years ago held roughly 730 to 820 milliliters (Spoor et al. 2015). One limit applies throughout: a skull shows how big a brain was, not what it was like to think with it.
Who were the earlier hominins?
“Hominin” means humans and our extinct relatives: everyone on our side of the family tree after it split from the chimpanzee line. Genetic estimates for that split range from about 4–6 million years ago to “at least 7–8 million years,” based on the mutation rates and generation times researchers assume (Langergraber et al. 2012). (Disputed: the exact date.)
Every fossil on this page comes from Africa, and not from one spot: Chad, Ethiopia, Kenya, Tanzania and South Africa. The fossil map also shows where people have looked: when Toumaï was announced in 2002, the search for early hominins had concentrated in East Africa (Brunet et al. 2002).
Who walked on two legs first?
Toumaï, Chad, about 7 million years ago. The oldest proposed hominin is a distorted but nearly complete skull nicknamed Toumaï, “hope of life” in the local Dazaga language (Smithsonian). Beryllium-isotope dating of the rock unit where the team places it gives 6.8 to 7.2 million years (Lebatard et al. 2008). A 2005 reconstruction estimated its braincase volume at 360–370 milliliters, within the chimpanzee range (Zollikofer et al. 2005, p. 758). (Plausible: reconstructed brain volume.)
Was it a hominin, and did it walk upright? (Disputed.) Critics objected in print within three months (Wolpoff et al. 2002); a 2005 digital rebuild argued for a hominin that might have walked upright (Zollikofer et al. 2005). A thigh bone found with the skull in 2001 was first taken for some other animal. Studies of it have flipped back and forth: no clear sign of two-legged walking in 2020 (as summarized by Williams et al. 2026), habitual two-legged walking plus climbing in a 2022 analysis (as described by Williams et al. 2026), no again in 2024 (Max Planck Society 2024), and in January 2026 a ligament attachment the authors say only bipedal hominins have (Williams et al. 2026). That last paper calls the recent research “polarized.”
Ardipithecus, Ethiopia, about 5.8 to 4.4 million years ago. Yohannes Haile-Selassie found the first Ardipithecus kadabba fossil in 1997; the species, about 5.8 to 5.2 million years old, is known from teeth and a few bones (Smithsonian). Tim White, Gen Suwa and Berhane Asfaw named the 4.4-million-year-old ramidus species in 1994 (White et al. 1994). The partial skeleton “Ardi,” found that year and announced in 2009, had a brain of about 300–350 milliliters and a grasping big toe for climbing (Smithsonian). Whether her crushed pelvis shows two-legged walking is questioned (Smithsonian). (Disputed.)
Ardi’s species had a surprise in its teeth: male and female canines differed in size less than in bonobos, the least size-different living great ape. The researchers read this as low male-to-male aggression, “most likely” shaped by female choice (Suwa et al. 2021). (Plausible: behavior inferred from teeth.)
Lucy and Laetoli, East Africa, about 3.85 to 2.95 million years ago. A. afarensis lived in what are now Ethiopia, Kenya and Tanzania, ate mainly plants, and probably ate the occasional small animal (Smithsonian). (Plausible: the small-animal part.) Males averaged about 151 centimeters tall, females about 105 (Smithsonian). Donald Johanson and Tom Gray found “Lucy,” a 3.2-million-year-old partial skeleton, at Hadar, Ethiopia, in 1974 (ASU Institute of Human Origins). Her pelvis and thigh bones indicate upright walking (Smithsonian).
Mary Leakey and Richard Hay described the Laetoli footprints in the journal Nature in 1979 (Leakey and Hay 1979). Most researchers credit the main trackways to A. afarensis (McNutt et al. 2021). (Plausible: footprints carry no name tag.) One experiment suggests the walkers used straight-legged, human-like strides (Raichlen et al. 2010), and a second, re-excavated trackway suggests at least two kinds of two-legged hominin lived there (McNutt et al. 2021). (Plausible: one study each.)
How big were their brains?
| Who | When | Brain estimate | How sure |
|---|---|---|---|
| Sahelanthropus (Toumaï) | ~7 million years ago | ~360–370 ml, within the chimpanzee range | Plausible |
| Ardipithecus ramidus (Ardi) | ~4.4 million years ago | ~300–350 ml | Plausible |
| Australopithecus afarensis | ~3.85–2.95 million years ago | average ~445 ml (range 365–526); Lucy ~388 | Strongly supported |
| Chimpanzees today | now | average ~369 ml | Documented |
| Homo habilis, KNM-ER 1813 | ~1.9 million years ago | ~510 ml | Plausible (single institutional estimate) |
| Homo habilis, OH 7 | ~1.8 million years ago | ~729–824 ml (combined uncertainty 630–967) | Plausible (one reconstruction) |
Table sources: Zollikofer et al. 2005, Smithsonian, Ardi, Gunz et al. 2020, Smithsonian, KNM-ER 1813, Spoor et al. 2015.
Analogy: a 12-ounce soda can holds 355 milliliters, close to the chimpanzee average. A half-liter water bottle (500 ml) holds a bit more than the A. afarensis average.
Across the family tree, brain size rose gradually, mostly within lineages (Du et al. 2018), and it varied a lot within each early Homo species (Spoor et al. 2015).
Can a skull tell us about a mind?
Brains decay; bone lasts. So researchers measure the space inside the braincase, a reliable stand-in for brain size (Du et al. 2018). Sometimes nature makes a cast of that space, called an endocast. The skull of the roughly 3.3-million-year-old Dikika child (“Selam”), found in Ethiopia by a team led by Zeresenay Alemseged, is filled with fine-grained sandstone, which forms a natural endocast (Smithsonian; Gunz et al. 2020).
Analogy: an endocast is like the dent your head leaves in a pillow. It shows size and rough shape, not what you were dreaming.
Using high-powered X-ray scans, Philipp Gunz and colleagues found the imprint of a key brain groove in an ape-like position on one adult skull. Comparing two infants with adults, they also concluded that A. afarensis brains took longer than chimpanzee brains to reach adult size. They suggested two explanations: a long childhood dependent on caregivers, or a way to spread out the energy cost of raising young (Gunz et al. 2020). (Plausible: two infant fossils; neither explanation established.)
A fossil can document size and some surface shape, and support careful inferences about growth. It cannot tell us what these individuals thought, felt, believed or communicated. (Unknown.) And a smaller brain is not proof of a simple mind.
Who made the first stone tools?
In 1964 Louis Leakey, P. V. Tobias and John Napier named Homo habilis, “handy man,” from fossils found at Olduvai Gorge, Tanzania (Leakey, Tobias and Napier 1964; Smithsonian). Scholars long assumed the earliest stone tools were made by Homo (Harmand et al. 2015). Newer finds complicate that:
- Lomekwi 3, Kenya, about 3.3 million years ago. Sonia Harmand, Jason Lewis and colleagues reported flaked and battered stones and named them the Lomekwian (Harmand et al. 2015). Critics say the tools were not convincingly shown to sit in their original layer (Domínguez-Rodrigo and Alcalá 2016). (Disputed.) The same area yielded Kenyanthropus platyops, a flat-faced, roughly 3.5-million-year-old skull named by Meave Leakey’s team in 2001 (Leakey et al. 2001); whether it is a separate genus is Disputed (Smithsonian). Who made the tools is Unknown.
- Nyayanga, Kenya, about 3 million years ago. Oldowan tools share a layer with cut-marked hippo bones and two teeth of Paranthropus, a different hominin; no Homo habilis fossils are known from the dig. Ruling out Paranthropus as a toolmaker is “unwise,” the Smithsonian’s program says (Smithsonian 2023).
- Earliest Homo, about 2.8 million years ago. A jaw from Ledi-Geraru, Ethiopia, provides evidence of Homo at about 2.8 million years ago (Harmand et al. 2015, note added in proof). Later finds there led the team to conclude that as many as four hominin lineages lived in eastern Africa between 3.0 and 2.5 million years ago (Villmoare et al. 2025).
So stone tools appear before the clear rise in brain size. (Strongly supported.) H. habilis itself, about 2.4 to 1.4 million years ago, ate a flexible diet that included some meat and marrow (Smithsonian).
Why did brains stay small so long, and then grow?
(Unknown.) Three leading ideas could each be partly right:
- Energy. Brains are costly. Adjusting for body size and activity, living humans burn roughly 400 more calories a day than chimpanzees and bonobos, enough to cover a bigger brain (Pontzer et al. 2016). Applying this to ancient diets is Plausible.
- Social life. Across living primates, the size of the neocortex (part of the cerebral cortex) tracks social complexity, and a bigger brain needs changes in diet and life history (how fast animals grow up and reproduce) to pay for it (Dunbar and Shultz 2007). Applying this to early hominins is Plausible; no fossil records their group sizes.
- Tools and learning. In great apes, evidence runs both ways: traditions need smart animals, and traditions make animals smarter (Whiten and van Schaik 2007). Early tools show toolmaking did not need a Homo-sized brain; whether tools later drove brain growth is Unknown.
How do we know any of this?
What evidence exists. Crushed skulls rebuilt by computer, endocasts, limb bones, footprints, stone tools, cut-marked animal bones and dated rock layers.
Whose voices are missing. None of these prehistoric individuals left a written account, so every statement about their minds is inference. Accounts also tend to shrink the many people who found fossils into “a team led by” one leader. Raymond Dart’s 1925 report on South Africa’s Taung child names the student and the owner behind an earlier Taung fossil, a monkey skull, but credits its blasting only to “operatives of the Northern Lime Company” (Dart 1925). The Smithsonian still lists Dart as the Taung child’s finder “via local quarrymen” (Smithsonian). Researchers including Kamoya Kimeu, who found KNM-ER 1813 (Smithsonian, KNM-ER 1813), Yohannes Haile-Selassie and Zeresenay Alemseged found or led many of the finds on this page. Researchers at Chadian institutions, including Hassane Taisso Mackaye and Andossa Likius, co-authored the 2002 Toumaï paper (Brunet et al. 2002).
How power shaped the record. Access and publishing speed shape what we know. Ardi was found in 1994 and announced in 2009. The thigh bone found near Toumaï waited until 2020 for its first published analysis.
Then vs. now
None of these prehistoric individuals left a written record of their beliefs, so this compares scientists’ earlier ideas with today’s evidence.
- Then: stone tools were long assumed to be a Homo invention (Harmand et al. 2015). Now: Oldowan tools date to about 3 million years ago beside Paranthropus teeth, and the older Lomekwian is disputed.
- Now: the fossil record supports several lineages living at once, including as many as four in eastern Africa between 3.0 and 2.5 million years ago (Villmoare et al. 2025). (Strongly supported as the authors’ interpretation.)
- Then: slow brain growth was seen as a side effect of big Homo brains. Now: A. afarensis suggests it started earlier (Gunz et al. 2020). (Plausible.)
Analogy: the old picture looked like a ladder. The evidence now looks more like a bush with several branches alive at once.
What remains unknown?
- Whether Sahelanthropus walked upright or belongs on the human line.
- Who made the earliest stone tools.
- Why brains grew when they did.
- What any of these individuals thought, felt or communicated. Fossils cannot answer that. (Unknown.)
