A new analysis of 2,200-year-old fingerprints pressed into the clay of China's Terracotta Army indicates the figures were not shaped by adult men alone. Researchers estimate that women made up about a third of the print-makers studied and children under 15 close to a quarter, alongside roughly 45% adult men.
China's Terracotta Army has been studied for half a century, but a question that has long puzzled archaeologists is who actually shaped the thousands of clay soldiers buried with Emperor Qin Shi Huang. A study published in the Journal of Cultural Heritage in 2026 and reported on by several outlets in mid-October offers a new answer: fingerprints left in the wet clay before firing.
Why this is back in the news
The study, led by researcher Xiaoxi Li of the Mausoleum Site Museum in Xi'an together with colleagues from Shandong University, appeared in the Journal of Cultural Heritage. Media coverage followed in mid-October 2026, more than five decades after the first terracotta figures were unearthed near Xi'an in 1974, according to World History Encyclopedia.
What the fingerprints show
The team examined 79 well-preserved fingerprints found on 36 figures from Unit T23 in Pit 1, a section that holds 227 figures, roughly a tenth of those excavated from that pit, according to Archaeology Magazine. Each figure yielded between one and five usable prints, mostly on the arms, armor and heads.
Instead of trying to classify each print individually, the researchers used a statistical technique called a Bayesian mixture model. It compares the mean ridge breadth of the ancient prints, the average spacing between fingerprint ridges, against reference data from living people of different ages and sexes, since ridge spacing tends to widen as a person grows. The team also factored in an estimated 2 to 12 percent shrinkage the clay would have undergone during drying and firing.
Based on that model, the study estimates that about 45% of the fingerprints came from adult men, roughly 32.7% from adult women and about 22.3% from children under the age of 15. Archaeology Magazine, which reported on the findings, noted the typical adult print-makers were young, around 20 years old for men and 18 for women.
Estimated origin of the 79 fingerprints studied
Group
Share of fingerprints
Typical age (where estimated)
Adult men
~45%
around 20
Adult women
~32.7%
around 18
Children under 15
~22.3%
not specified
The artisan community was not exclusively composed of adult males.
— Study authors, Journal of Cultural Heritage
The researchers suggested, according to Archaeology Magazine, that women and younger workers may have carried out tasks such as smoothing clay surfaces or attaching armor details, often working alongside more experienced craftsmen rather than producing entire figures alone.
Convicts and forced labor: what historical records already said
The fingerprint findings add to, rather than replace, long-standing evidence about who worked on Emperor Qin Shi Huang's funerary complex. Historian Sima Qian's ancient account, cited by World History Encyclopedia, describes more than 700,000 convict laborers sent to work on the wider mausoleum project.
Physical evidence of forced labor has also surfaced at Zhengzhuang, near the mausoleum wall, where archaeologists found iron neck collars and leg shackles, according to Archaeology Magazine. That site is separate from the Terracotta Army pits examined in the new fingerprint study.
Background: what the Terracotta Army is
The Terracotta Army was created in the third century BC as part of the funerary complex of Qin Shi Huang, China's first emperor, who died in 210 BCE. According to World History Encyclopedia, construction of the broader mausoleum began early in his reign, and the army itself comprises more than 7,000 warriors, 600 horses and 100 chariots.
The site, located near Xi'an in Shaanxi province, was discovered by farmers digging a well on 29 March 1974.
What the new study does not prove
The researchers themselves caution that their sample is limited: 79 fingerprints from a single unit cannot be assumed to represent the demographics of every workshop that produced the thousands of figures across all pits, a point reported by both Phys.org and HeritageDaily. The model also relies partly on modern reference fingerprint data, and ridge-spacing patterns can vary somewhat between populations, meaning the age and sex estimates carry statistical uncertainty rather than certainty.
79 fingerprints from 36 figures in Unit T23, Pit 1, were analysed.
Estimated composition: about 45% adult men, 32.7% adult women, 22.3% children under 15.
Historical texts already documented convict labor on the wider mausoleum project.
Iron shackles found at nearby Zhengzhuang confirm forced labor was used at the site, separate from the pits studied here.
The authors describe the findings as estimates from one sample, not a full workforce census.
Questions
Who built China's Terracotta Army, according to the new study?
The fingerprint study estimates that adult men made up about 45% of the print-makers, adult women about 32.7%, and children under 15 about 22.3%, based on 79 fingerprints from 36 figures in one excavation unit.
How did researchers determine the age and sex of the workers from fingerprints?
They measured the mean ridge breadth, the average spacing between fingerprint ridges, which widens as a person ages, and used a Bayesian mixture model to compare the ancient prints with modern reference data, while accounting for clay shrinkage during firing.
Were convicts used to build the Terracotta Army?
Historical records cited by World History Encyclopedia describe over 700,000 convict laborers working on Qin Shi Huang's broader mausoleum project, and archaeologists have found iron shackles at the nearby Zhengzhuang site, evidence of forced labor, though separate from the specific pit examined in the fingerprint study.
How many warriors are in the Terracotta Army and when was it discovered?
The army includes more than 7,000 warriors, 600 horses and 100 chariots, and was discovered by farmers near Xi'an on 29 March 1974.
Does this study prove most of the Terracotta Army was made by women and children?
No. The study's authors and outlets reporting on it note the sample of 79 fingerprints comes from a single unit and should not be generalized to the entire workforce across all pits.
This piece was prepared with artificial intelligence and passed an automated source check before publication. If you see an error, ask for a correction. Request a correction or reply · How we make this
Photosynthesis is the process by which plants, algae and some bacteria use sunlight, water and carbon dioxide to make sugar and oxygen. It happens in two linked stages inside chloroplasts: light-dependent reactions that capture solar energy, and the Calvin cycle that uses that energy to build carbohydrates.
Vaccines work by showing the immune system a harmless piece of a pathogen — called an antigen — so the body builds antibodies and memory cells without suffering the actual disease. Getting a vaccine to that point takes years of preclinical testing and three phases of human trials before regulators such as the FDA license it. Full protection after vaccination typically takes a few weeks, as the immune system needs time to produce lasting defenses.
Warhound, co-founded by Colossal Biosciences' Ben Lamm and Anduril's Palmer Luckey, raised $200 million on October 22, 2026 to genetically edit and clone dogs for the US military, aiming to remove inherited diseases from breeds such as the Belgian Malinois. Independent scientists call the genetic goals technically achievable, but bioethicists warn that cloning remains inefficient and raises unresolved animal-welfare and regulatory questions.
Leap seconds, the one-second adjustments used since 1972 to keep clocks aligned with Earth's rotation, will stop entirely. The world's timekeeping authority, the International Bureau of Weights and Measures (BIPM), confirmed on 15 October 2026 that continuous UTC takes effect on 20 May 2027. The move, first agreed in principle in 2022, was accelerated after experts flagged a growing risk of a disruptive "negative" leap second.
