The plaster casts of Pompeii are among the most affecting objects in archaeology. When Vesuvius buried the town in 79 CE, ash compacted around the bodies of the dying; the soft tissue decayed and left cavities, and nineteenth-century excavators filled those voids with plaster, recovering the outlines of people in their final moments.
Around those shapes, stories accumulated. A figure wearing a golden bracelet with a child on their lap became a mother and her child. Two figures who appeared to have died together became sisters. Those readings entered guidebooks, museum labels and a century of popular writing.
A 2024 study tested five of them against DNA. The stories did not survive.
The evidence comes from Ancient DNA challenges prevailing interpretations of the Pompeii plaster casts in Current Biology, which generated genome-wide data and strontium isotope measurements from skeletal material embedded in the casts of five individuals.
The short answer: the sexes and family relationships of the sampled individuals do not match the traditional interpretations. The adult wearing a golden bracelet with a child on their lap was a genetic male, biologically unrelated to the child. The pair long read as sisters included at least one genetic male. All the sampled Pompeiians derive their ancestry largely from recent immigrants from the eastern Mediterranean — matching contemporaneous genomes from Rome. These are five individuals, not the population of a town.
How the casts complicate the science#
The casts are not skeletons. They are plaster poured into cavities, sometimes over bones and sometimes not, produced by a technique developed in the 1860s and applied repeatedly since. Later restoration work introduced additional material, and the casts have been handled, displayed and moved for a century and a half.
That history creates real difficulties for ancient DNA: contamination from everyone who worked on them, plaster physically bound to bone, and fragmentary skeletal remains that were never excavated with genetic sampling in mind. The study worked with material from 14 of the 86 casts undergoing restoration, and obtained genome-wide data from five.
Two independent measurements were taken from that material:
| Measurement | What it establishes |
|---|---|
| Genome-wide ancient DNA (>1 million SNP targets) | Genetic sex, biological relatedness, genome-wide ancestry |
| Mitochondrial DNA enrichment | Maternal lineage |
| Strontium isotopes | Whether a person grew up locally or moved during childhood |
Genetic sex determination is among the most robust results ancient DNA produces — it depends on the ratio of reads mapping to the sex chromosomes, and it works at coverage far too low for most other analyses.
The bracelet, the child, and the assumption#
The best-known correction concerns the cast of an adult wearing a golden bracelet with a child on their lap. Read as a mother sheltering her child, it has been reproduced in that framing for generations.
The adult was genetically male, and was biologically unrelated to the child.
The pair often described as sisters, who appeared to have died in an embrace, included at least one genetic male.
The authors are careful about what follows. The corrections do not establish who these people were — a father, an uncle, an enslaved person, a guardian, a stranger caught in the same doorway are all consistent with the data. What the results demonstrate is that the original readings came from modern assumptions about gendered behaviour rather than from evidence: an adult comforting a child was assumed to be a mother; two people embracing were assumed to be women.
That is the paper's own stated lesson, and it is stated in its title. Modern assumptions about gender are not a reliable lens through which to view the past.
A cosmopolitan town, as expected#
The ancestry results are less dramatic but more historically significant. All the Pompeiians with genome-wide data derive their ancestry largely from recent immigrants from the eastern Mediterranean.
That matches what ancient DNA from the city of Rome in the same period had already shown. The Roman Imperial period saw substantial movement of people across the Mediterranean — through trade, military service, administration, and the slave trade, which forcibly transported enormous numbers of people from the eastern provinces into Italy.
The consequence is that "Roman" in the first century CE was a political and cultural category, not an ancestry. A person born in Pompeii to parents from the eastern Mediterranean was a Pompeiian, and possibly a Roman citizen, with no contradiction. Similar patterns appear across the empire's provinces — the Balkan frontier and the central European limes after Rome both show the same mobility, in different directions.
Strontium isotopes add an individual-scale dimension. The ratio of strontium isotopes in tooth enamel reflects the geology of where a person spent childhood, so it can distinguish someone raised locally from someone who arrived later — a distinction genome-wide ancestry alone cannot make, since a person of eastern Mediterranean ancestry may have been born and raised in Pompeii.
Five people are not a town#
The most important caveat is one of scale. Pompeii's population at the time of the eruption is usually estimated in the low tens of thousands. This study reports genome-wide data from five individuals.
Five people cannot describe a town's demography, and they were not randomly selected. They come from the subset of casts currently under restoration, which is itself a subset of the casts that were made, which reflects where excavation happened and which cavities were noticed and filled. Every stage of that chain is a selection.
So the ancestry finding should be read as: these five individuals had substantial eastern Mediterranean ancestry, consistent with independent evidence from Rome that such ancestry was common in Roman Italy. It is not a measurement of what proportion of Pompeiians did.
The kinship corrections are on firmer ground, because they are claims about specific individuals rather than about a population — and specific individuals is exactly what the traditional interpretations were about.
What genetic sex determination does and does not say#
Because the corrections here turn on sex, it is worth being exact about what was measured.
Genetic sex is read from the proportion of sequencing reads that map to the X and Y chromosomes. An individual with one X and one Y produces a characteristic ratio; an individual with two X chromosomes produces another. It requires no intact gene, works at low coverage, and is among the first things any ancient-DNA pipeline reports.
What it establishes is chromosomal sex. It does not establish gender, social role or how a person lived and was regarded — categories that are cultural, varied historically, and inaccessible to a genome. A first-century Roman household contained relationships that modern categories map onto poorly, including enslaved people whose position in a household was not kinship in any modern sense but who lived, worked and died inside it.
So the finding is that the adult holding the child was chromosomally male and not the child's biological relative. The finding is not that the pair had no relationship, or that the scene means less than it appeared to. It means the specific story attached to it — mother and child — was an assumption, and the assumption was wrong.
Why the original stories were told#
It would be unfair to treat the nineteenth-century interpretations as simple carelessness. Giuseppe Fiorelli's casting technique was a genuine innovation, and the excavators had almost nothing but posture and position to work with. In the absence of other evidence, a narrative reading was the only reading available.
The problem is what happened next: readings offered as plausible were repeated until they hardened into fact, and the uncertainty was lost somewhere between the excavation report and the museum label. By the time anyone could test them, they were not presented as hypotheses at all.
That pattern is not confined to Pompeii. It is the same mechanism that produced the "mother and child" burials, "warrior" graves and "princess" burials across European archaeology that ancient DNA has been steadily correcting — many of which were sexed from grave goods rather than from bone.
Limitations#
| Limitation | Why it matters |
|---|---|
| Five individuals with genome-wide data | Too few to characterise Pompeii's population. |
| The sample is not random | It reflects which casts were made, survived and are under restoration. |
| Casts are difficult material | Plaster, restoration and handling complicate recovery and raise contamination risk. |
| Corrections are negative results | They show what a relationship was not, rarely what it was. |
| Ancestry is not identity | Eastern Mediterranean ancestry says nothing about how a person self-identified. |
| Isotopes indicate region, not origin | Strontium narrows childhood geology; it does not name a birthplace. |
Frequently asked questions about the Pompeii plaster-cast DNA#
How many individuals were studied?#
Genome-wide ancient DNA and strontium isotope data were generated for five individuals, drawn from work on 14 of the 86 casts undergoing restoration.
What was wrong with the "mother and child" cast?#
The adult wearing a golden bracelet with a child on their lap was genetically male and biologically unrelated to the child. The relationship between them is unknown.
And the pair thought to be sisters?#
At least one of the two individuals long interpreted as sisters was genetically male.
Where did the Pompeiians come from?#
The sampled individuals derive their ancestry largely from recent immigrants from the eastern Mediterranean, consistent with contemporaneous genomes from the city of Rome and with the mobility of the Roman Imperial period.
Can DNA be recovered from the plaster itself?#
No. The plaster is a nineteenth-century material. The DNA comes from skeletal remains embedded within the casts, which is why only some casts are viable at all.
Does this change what the casts mean?#
It changes the stories told about specific casts, not their significance. They remain a direct record of individual people at the moment of a catastrophe — the corrections concern what later observers assumed about who those people were to one another.
Evidence against a good story#
The Pompeii casts are effective precisely because they invite narrative. A person crouching, an adult holding a child, two figures side by side: the shapes ask to be read, and for a century and a half they were.
What this study supplies is the first independent check. In every case tested, the traditional reading was wrong — not marginally, but on the basic facts of sex and relatedness. The individuals remain as affecting as they were; what has gone is the confidence that we knew who they were.
The broader value is methodological. Archaeological interpretation has always filled gaps with plausibility, and plausibility is shaped by the assumptions of whoever is doing the filling. Genome-wide data does not eliminate interpretation, but it does supply facts that can contradict it — and, in this case, did so five times out of five.
Sources and further reading#
- Pilli, E., Vai, S., Moses, V. C. et al. (2024). Ancient DNA challenges prevailing interpretations of the Pompeii plaster casts. Current Biology 34, 5307–5318.e7. DOI: 10.1016/j.cub.2024.10.007.
- Antonio, M. L., Gao, Z., Moots, H. M. et al. (2019). Ancient Rome: A genetic crossroads of Europe and the Mediterranean. Science 366, 708–714.
- Parco Archeologico di Pompei, conservation and restoration reports on the plaster casts: pompeiisites.org.
Editorial note: this article was written as a source-based synthesis and distinguishes genetic relatedness from social relationships throughout. Its hero and section artwork was generated with AI as an interpretive architectural scene, not as scientific evidence, and does not depict the casts or the individuals preserved in them.
