1-Million-Year-Old Skull Found in China Could Rewrite Human Evolution

Ancient find – In 1990 researchers uncovered a badly crushed human skull (now called Yunxian 2) in Hubei province, central China. Recent 3D CT-scanning and digital reconstruction have revealed it is about 940,000–1.1 million years old ^[1]. This reanalysis shows it does not match classic Homo erectus, but instead belongs to a sister lineage linked to Denisovans and “Dragon Man” (Homo longi) ^[2] ^[3].
Timeline upheaval – If confirmed, Yunxian 2 implies our own lineage diverged from other hominins far earlier than thought. The authors suggest that by ~1 million years ago the various human lineages had already split into distinct branches ^[4] ^[5]. In effect, modern humans (H. sapiens) might have emerged ~400,000 years earlier than the fossil record currently shows ^[6] ^[7].
Expert reaction – Paleontologist Chris Stringer (Natural History Museum, London) says “this changes a lot of thinking, because it suggests that by one million years ago, our ancestors had already split into distinct groups” ^[8]. Lead author Xijun Ni (Fudan Univ., Shanghai) adds they “challenged the old timelines of human evolution” ^[9]. However, others urge caution: Andy Herries (La Trobe Univ.) notes this is a remarkable fossil but warns the analysis “was based on two different skulls” and “trying to cram it into H. longi is… unwarranted given all the uncertainties” ^[10] ^[11].
Dragon Man link – The finding ties into other recent discoveries in Asia. Just weeks earlier (June 2025), scientists announced that the famous Harbin “Dragon Man” skull (146,000 years old) is a Denisovan based on ancient DNA and protein evidence ^[12]. In other words, Dragon Man and Yunxian 2 likely belong to the same broad Asian hominin branch. As a Science journal summary noted, “Denisovans looked like Dragon Man” ^[13] – and now Yunxian 2 appears on that same tree.
New picture of the Middle Pleistocene – This discovery helps resolve the so-called “muddle in the middle” of human evolution (the 300,000–1,000,000-year period where fossils are scarce and hard to classify). The researchers propose five major clades of Homo (including sapiens, Neanderthals, Denisovan/longi, heidelbergensis, erectus) diverged over a million years ago ^[14]. Yunxian 2 lies near the origin of the Denisovan/H. longi line and provides a transitional snapshot. As Stringer puts it, if it “sits close to the origins” of both the Denisovan and sapiens lines, it may be “one of the most important windows yet into the evolutionary processes that shaped our genus” ^[15].

These key points set the stage for a detailed exploration of the new study and its implications. In the sections below we explain the discovery and reconstruction of the skull, compare it with other recent finds, and examine what experts are saying about how it might upend our understanding of human origins.

The Yunxian 2 Skull: Discovery and Digital Reconstruction

The fossil at the heart of this story is the Yunxian 2 cranium, unearthed in 1990 from a river terrace in Yunxian (Hubei). For decades the specimen sat in a museum, considered an indeterminate Homo erectus fossil because it was badly crushed. Only now have high-resolution CT scans and 3D modeling unraveled its shape ^[16] ^[17]. By virtually “de-warping” the skull, researchers could examine its original anatomy. The result is a surprisingly modern-looking face on a very ancient skull.

According to Reuters, the digital reconstruction shows Yunxian 2 is a male in his 30s and about 940,000–1.1 million years old ^[18]. Importantly, it is not a textbook Homo erectus. Instead, its combination of features is a “mosaic” of primitive and advanced traits: a large braincase (the largest measured for any human of that age), flat and forward-facing cheekbones, a projecting nose on a wide face, and thick browridges, but no Neanderthal-like midface bulge ^[19]. As Stringer describes it, “It has a long, low skull and receding forehead behind a strong browridge… [and] a large nose with a projecting nasal bridge, but without the midfacial prominence we find in Neanderthals” ^[20]. Xijun Ni adds that the skull shares with the Denisovan/longi group “a broad and massive roof of the mouth, flat and low cheekbones, an expanded region at the back of the head and some features of the ear region” ^[21].

In short, detailed analysis clearly places Yunxian 2 in an Asia-centered hominin lineage. This line includes the recently named Homo longi (“Dragon Man”) and the Denisovans. In particular, statistical and phylogenetic tests find Yunxian 2 groups more closely with the Homo longi/Denisovan clade than with any African or European hominin. The Archaeology Magazine summary notes it looks like a “transitional form between Homo erectus and later groups,” sister to the sapiens line ^[22]. In practical terms, the skull’s anatomy suggests that around 1 million years ago there were at least two broad human lineages in Eurasia – one leading ultimately to us, and one to the Denisovan/Dragon-Man branch ^[23] ^[24].

A Timeline Rewritten: How Far Back Does Homo Sapiens Go?

The real bombshell of the new study is how old the human family tree may be. Most genetic clocks and fossils have pointed to modern Homo sapiens splitting from other human branches around 400,000–600,000 years ago. (For example, Neanderthals and humans were thought to separate about 600,000 years ago.) But Yunxian 2 forces a rethink. Because it is about 1 million years old and already looks “derived” (more advanced) than classic erectus, it implies that the divergence of lineages happened much earlier.

The Science paper’s authors propose that five major Homo clades diverged over a million years ago ^[25]. In their model, Neanderthals split off ~1.38 million years ago, then the Denisovan/Homo longi clade parted ways with the future human (sapiens) line by ~1.32 million years ago ^[26]. Notably, “early Homo sapiens traits” begin to appear around 1.02 million years ago, well before the roughly 300,000-year-old African fossils traditionally called Homo sapiens. Stringer emphasizes that Yunxian 2 sits near that divergence: if it is an early member of the longi/Denisovan clade, then by one million years ago our lineage had already been separate. In his words, “This changes a lot of thinking” – implying Homo sapiens may have emerged several hundred thousand years earlier than the current textbook date ^[27].

Ni and colleagues conclude: “Based on our new discovery, we challenged the old timelines of human evolution,” arguing that the human branch (Homo sapiens) likely split from Asian cousins long before previously thought ^[28]. In effect, they suggest our species’ origins began around a million years ago, not just 300–500 thousand years ago. As the CNN-reported summary explains, if widely accepted this would push back humanity’s start by roughly 400,000 years ^[29]. Such an adjustment would have ripple effects – altering estimates for when cultural or genetic traits arose.

The study also ties into climate context. Stringer points out that two severe ice-age cold snaps occurred around 1.1 and 0.9 million years ago ^[30]. These extreme environmental pressures could have driven rapid evolution and even local extinctions, plausibly fragmenting populations and spurring the divergence seen in the fossil record. In short, rather than a slow, linear progression, the new evidence favors a complex branching of multiple human groups in the Middle Pleistocene.

Experts Weigh In: Support and Skepticism

Support: Leading paleoanthropologists find the study provocative. Chris Stringer calls Yunxian 2 a potential “window” into early human evolution ^[31]. He explains that in the 1970s and ’80s the Asian fossil record was often ignored or seen as peripheral; now fossils like Yunxian, Harbin (Dragon Man) and Denisova force a new view of Asia as a crucial region for our story ^[32] ^[33]. Ni points out that the Homo longi/Denisovan clade “was quite successful in Asia, occupying a very large area with diverse environments for more than one million years” ^[34]. In other words, a diverse Denisovan-related group thrived in Asia for a very long time, largely isolated from African/European lineages.

In media comments, Stringer said that if Yunxian 2 really sits near the base of both the Denisovan and sapiens lines, then it “may represent one of the most important windows yet into the evolutionary processes that shaped our genus” ^[35]. The ABC Science article quotes him noting that this find helps resolve the puzzling “Muddle in the Middle” – the period (roughly 300k–1M years ago) when many archaic human fossils defied easy classification ^[36] ^[37]. By providing a clear example of an early Asian Homo, Yunxian 2 could anchor many of those mystery fossils.

Skepticism: Other researchers urge caution. Professor Andy Herries (La Trobe University, Australia) points out uncertainties: for instance, the digital reconstruction of Yunxian 2 actually used fragments from two different skulls, which may bias the result ^[38]. He agrees the fossil is important, but calls the H. longi/Denisovan assignment “unwarranted given all the uncertainties” ^[39]. Geneticist María Martinón-Torres (Nature correspondence) notes that without DNA from the fossil, the conclusions rely on morphology only – and convergent features can sometimes mislead. The researchers themselves acknowledge the need for further work, and they state that direct dating of the site or additional finds would bolster their case.

In summary, experts are excited that Asia’s fossil record is finally shedding light on our roots, but many agree that extraordinary claims demand extensive verification. As one commentator puts it, the results will need “time to be reviewed and challenged” ^[40]. For now, these new findings are a major headline – but a few too early to rewrite textbooks without further corroboration.

The Wider Context: Asian Hominins and Human Origins

The Yunxian 2 discovery is one piece in a larger puzzle that has been shifting rapidly in the past few years. Most notably, it dovetails with the June 2025 announcement that the Harbin “Dragon Man” (Homo longi) is actually a Denisovan. Two new studies (Science and Cell) extracted ancient DNA and proteins from the 146k-year-old Harbin skull and found a clear Denisovan signature ^[41]. In their words, researchers “have proved that Dragon Man is indeed the face of Denisovans” ^[42], and that “Denisovans looked like Dragon Man” ^[43]. This confirms that Homo longi and the Denisovans are the same lineage – meaning that Yunxian 2, also on that lineage, is a direct part of the Denisovan story.

Aside from DNA, the Denisovans had been known only from small Siberian fossils and genetic traces (e.g. modern Asian and Pacific peoples carry Denisovan DNA). Now with Harbin and Yunxian, scientists finally have credible skulls to match the DNA – a first morphological “blueprint” for Denisovans ^[44] ^[45]. In fact, with Harbin (China), Denisova Cave (Russia), and even a tooth found off Taiwan (also identified as Denisovan in April 2025), “paleoanthropologists have definitive examples that other unknown skulls can be compared to” ^[46].

Yunxian 2 also fits with the idea that Asia was not an empty backwater of human evolution. Over the past decade, discoveries like Homo floresiensis (Indonesia), Homo luzonensis (Philippines), and archaic fossils in China and Southeast Asia have shown an array of human cousins coexisting. The new analysis suggests these lineages were not recent arrivals: rather, descendants of an early branching. In particular, it implies that the line leading to modern humans and the line leading to Denisovans split in Asia well before 1 million years ago, whereas previous models often assumed the divergence happened after Homo erectus had already appeared worldwide.

This feeds into the long-standing debate on human origins: Out-of-Africa versus multiregional evolution. The established view – supported by genetics – is that Homo sapiens arose in Africa (earliest fossils ~315k years old in Morocco) and later spread worldwide, replacing other hominins. But the complex Asian record invites a twist: some scientists argue that early Asian populations (like those represented by Yunxian and Harbin) may have contributed to the gene pool of later humans, or at least influenced our evolution through deep coevolution. New finds like this do not overthrow Out-of-Africa, but they do highlight that multiple human species “branched” early on, and Asia’s hominins were diverse and long-lived.

Conclusion: Rewriting the Human Story?

The study of the 1-million-year-old Chinese skull has opened a provocative new chapter in human paleontology. By combining modern imaging technology with comparative anatomy, researchers have placed an otherwise-forgotten fossil into a lineage that bridges Homo erectus and later hominins like the Denisovans and ourselves. If the conclusions hold, it suggests our own branch of the human tree split off far earlier than the classic timeline – possibly a full 1.0–1.4 million years ago. In the words of the scientists, Yunxian 2 “pushes back” the emergence of Homo sapiens and shows that the major human clades were already diverging around 1 million years ago ^[47] ^[48].

Experts find this work exciting, as it helps resolve the “muddle in the middle” of our fossil record. As Stringer notes, East Asia “preserves crucial clues to the later stages of human evolution in general” ^[49]. Yet they also emphasize that further evidence (new fossils, better dates, even ancient proteins) will be needed to fully rewrite human evolutionary history. For now, the Yunxian 2 skull stands as a remarkable and tantalizing hint that Asia played a key role in shaping our origins. It reminds us that the story of humanity is still being written – sometimes by the ghostly fragments of skulls lost and found, and now brought back to life by technology and teamwork.

Sources: Recent reports and studies in Science, Nature, CNN, BBC, Livescience, Reuters and others provide details of this discovery and its context ^[50] ^[51] ^[52] ^[53] ^[54] ^[55]. (See above for full references.)

Fossils 101 | National Geographic

Watch this video on YouTube.

References