Xu Fu’s eastward journey, taking 3,000 boys and girls, “vanished” for two millennia—yet the Japanese thump their chests: He’s our forefather. Decades of promotion, countless temples and inscriptions, even prime ministers paying ancestral homage.
But one round of DNA testing later, the results hit like a thunderclap—no Xu Fu, no Central Plains origin, ancestors from a wholly different bloodline.
Careful, or You’ll Capsize! “Xu Fu Ancestor Tale” Shot Down by DNA
Many Japanese love saying: “We’re Xu Fu’s descendants.” In certain settings, it sparks a real “cultural pride.” But a DNA report drops, and that facade crumbles.
In 2013, Jilin University teamed up with the University of Hong Kong and Peking University for a project: Extracting bones from Han, Tang, and Song tombs in Zhengzhou for full genetic, cranial, and anthropological analysis. Their aim was clear: Pin down the Han-era “Central Plains” genetic blueprint and see who it matches closest.
The findings were startling. Cranially, Japanese do resemble Han Central Plains folk—especially in nasal projection, orbital height, and zygomatic breadth—closest to modern Hong Kong Han traits. But shift to genetic makeup, and the story flips.
Mitochondrial DNA, tracing maternal lines with low mutation, was sampled from 2,500-year-old Qi state tombs in Linyi, Shandong. Results: Central Plains folk belong to haplogroup R’s B and F subclades.
Note: This “R system” appears in Europe too, but branches differ. Here, it’s Eastern lineages; Japanese do carry F groups, but scattered thinly, low frequency.
If Xu Fu truly brought 3,000 boys and girls to Japan, logic dictates a clear genetic imprint. Reality: Absent. F groups aren’t dominant in modern Japanese DNA, nor ancestral sources.
You say they look Han? Genes say no. Unlike? Faces overlap. What’s the deal?
Consensus now: Japanese stem from three ancestral waves—Jomon, Yayoi, Kofun peoples. Jomon from ~10,000 years ago, Amur River basin; Yayoi around 3rd century BCE, Korean Peninsula-bound; Kofun circa 3rd century CE, horse-riding with Turkic traits.
Japan’s bloodline is mixed, but stably so—no “Xu Fu” mix-in. Those claiming “our ancestors are Xu Fu’s Central Plains Han” might’ve swallowed too many tales and fancies.
Where Did the 3,000 Boys and Girls Drift To?
Qin Shi Huang’s second dispatch of Xu Fu eastward is logged in Records of the Grand Historian and Book of Han. Date: 219 BCE; launch: Langya, modern eastern Shandong coast.
First voyage: Xu Fu “volunteered”—seeking immortality elixir. Crew: Hefty—3,000 boys and girls, artisan specialists, fleet, three years’ provisions—like an “emigration squad.”
Years out, no word. By 210 BCE, Qin summons him back for elixir update. Xu Fu: “Mermaids blocked the way—can’t slay ’em, can’t get the drug.” Qin, oddly unflaring, sends a fleet to “clear mermaids.” Xu Fu sails again, vanishes for good.
Where to? Mainstream theory: Korea’s Pyeongtaek Gwangeo area. Qing scholar Wang Duo parsed Records‘ “reached plains of Gwangeo, never returned” as modern southwestern Korea.
Japanese folklore begs to differ: Xu Fu hit Kumano, Wakayama, farmed Fuji foothills, tutored “Emperor Jimmu,” spawned lines. Sounds fleshed-out, but details hollow.
Launch sites vary wildly: Some say Shandong’s Rongcheng; others Jiangsu’s Lianyungang Gan’gu Xu Fu Village; still others Qingdao’s Dongyujiahe Port. All sorta right, none precise; all sorta wrong, yet all boast steles, temples, museums.
Fun fact: Yoshiro Hayashi—Japanese ex-PM—hit Lianyungang’s Xu Fu Village in 2007 for ancestor rites. He swore by it: “I’m Xu Fu’s 80th-generation kin.” Chinese hosts obliged—cannons, banquets, flowers—a spectacle grander than Confucius homage.
But the rub: Did Xu Fu reach Japan? Direct proof? Sources say flat-out: No tomb inscriptions, no ancient texts, no remains, no DNA.
Genealogies? Plenty “Sudo” or “Fukushima” clans claim Xu Fu descent, toting “Fu” as blood badge. But lineages max out at Edo era; pre-that, blanks. Amusingly, early ones on wooden boards—temple fire, poof, gone.
In short, “Xu Fu to Japan” lacks verifiable “landfall”—just endless “imaginative sprawl.”
Faces Echo Central Plains, But Blood Denies Kinship?
Folks say Japanese faces mimic Han—spot on. Per Jilin University’s cranial study, facial traits align with Han Central Plains. But remember: Looks can align; genes don’t fib.
DNA’s long decoded Japanese origins. Simply: Triple mix—Jomon (~10,000 BCE, Northeast Asia south to Honshu/Hokkaido); Yayoi (3rd century BCE, Korean east-crossing, rice/bronze/iron bearers); Kofun (3rd-6th centuries CE, equestrian polity-shapers).
These three forged modern Japanese genetics. Which wave “Central Plains Han”? None.
Genes show Yayoi closer to Korean Peninsula, Jilin/Liaoning—not Guanzhong or Qi-Lu heartlands. Mitochondrial-wise, Japanese R-F subclade match <15%, dwarfing Hong Kong/Jiangnan Han rates.
Xu Fu’s 3,000? Recruited Central Plains/Qi-Lu/Guanzhong stock. If landed Japan, expect stark Central Plains markers today. Nope.
Stinger: Japanese genetically nearer Austronesian kin like Vietnamese/Philippinos than Central Han. So why cling to “Xu Fu descendants”?
It flatters. Grants an isle-nation “cultural roots” balm. “Ancestors from ancient civ” trumps “fisher-hunter forebears” in dignity.
Even “Xu Fu traces” in Japan? Mostly retrofitted. “Xu Fu Temples/Wells/Tombs”? Edo mid-late builds. Timeline from Xu Fu’s era: 1,700+ years gap.
If he spawned lines, 3,000 could’ve seeded an “intra-isle diffusion” clan. But ancient DNA/cluster archaeology? Zero “Xu Fu descendant group” traces.
Verdict clear: Japanese aren’t Xu Fu’s heirs. Just a millennium-old folktale, skewered by archaeology and genetics.
References
- Ancient Human Genetics Unveiled: Where Do the Japanese Ancestors Really Come From? – Phoenix Net, November 15, 2021
- Did Xu Fu Really Reach Japan? – China Culture Research Institute, July 5, 2019
- Are the Japanese Descendants of Xu Fu? – People’s Daily Online, April 22, 2014
- Xu Fu’s Eastward Voyage: Historical Mists Behind the Millennium Legend – Sohu, July 30, 2017
- Tracing the Xu Fu Eastward Voyage Legend – Yiqi Du Lishi, July 30, 2025
