Small attachments conventionally described as buttons occur in Greek and Roman dress. They should be distinguished from fibulae, mounts and strap fittings. On wide women's chitons, a series of joins along the upper edge of the cloth could form a sleeve-like section; their appearance does not always reveal how the fastening worked.
Museum collections include metal objects with domed heads and attachments on the reverse. The British Museum, for example, describes EA23883 from Tell Dafana as a copper-alloy button with a bar for sewing it on. Its attribution to the Roman period is marked with a question mark; that uncertainty should be retained.
Materials survive unevenly, but a selection of metal finds cannot establish their proportion in ancient clothing as a whole. A button-like shape does not by itself prove that an object fastened a garment rather than another item.
Ancient buttons could be flat, domed, or mushroom-shaped, and their method of attachment depended on material and use. Metal examples often had an eyelet, loop, or shank on the reverse. For reconstruction, diameter, height, profile, and traces of attachment are important because these features help distinguish a working fastener from a decorative mount.
Identification depends on the reverse, wear and the find context. A loop may allow sewing, but similar loops occur on pendants and equipment fittings. Not every round mount is a button.
The fastening of a paenula or tunic should follow a specific prototype. A modern row of buttons and cut buttonholes should not be added merely for convenience: ancient textile attachments could operate differently.
Record a specific find's dimensions, material, profile and reverse attachment before making a replica. Check how it connects to the cloth and whether it damages the fabric. Where evidence is incomplete, distinguish a proposed fastening method from the documented form of the object.
Material, finish and the number of attachments should follow the prototype. A modern button is not historically appropriate simply because it is small or plain.




Enamelled object identified as a button, with a thin triangular projection. Chester, England; second century CE.
Enamelled object identified as a Roman button. Metropolitan Museum of Art, New York; second–fourth century CE.
Object described as a bronze button in a private collection; reported date: first century BCE. Provenance, date and function are not established by the documentation provided.
Object described as a bronze button in a private collection; reported date: first century BCE. Provenance, date and function are not established by the documentation provided.Interested in Ancient Rome beyond reading? Join Legio X Fretensis or explore our reenactment directions.