Roman Name Generator

Names built on the tria nomina — the real three-part system Roman citizens used — with regular Latin endings. Invented rather than attested, so you get a plausible Roman who did not actually exist.

Built from a sound system, so the supply never runs out.

More options

Set the longest given name to 0 for no limit.

Names this produces

Generated when this page was built, using the same sound system the tool above uses.

  • Masculine, full form Vuclbequia Flavelius · Rafaius Antonicus · Cuvetu Fabinus · Crinaeo Flavinus · Gofmeanus Livicus · Vado Fabilius
  • Feminine, full form Gudaia Auricus · Cegqueprael Flavilius · Vomcia Valianus · Cotucaeella Marcicus · Siarfiailla Antonelia · Pimiaf Flavicus
  • Given names only Daemiia · Quosclup · Mialpraefa · Nimdeg · Muquius · Promes · Pocronven · Feprdeia · Quetiaella · Rotrquaquia · Pilisu · Poqua

The tria nomina, and why there were only about seventeen first names

Roman naming is the only tradition on this site that is fully documented, because the Romans wrote everything down. It is also the one that works least like a modern name, which is where fiction usually goes wrong.

A citizen man carried three names. The praenomen was personal. The nomen identified his gens, the broad clan. The cognomen began as a nickname and hardened into the name of a branch within that clan. Gaius Julius Caesar is the standard illustration: Gaius personal, Julius for the gens Julia, Caesar for his branch of it.

The striking part is the praenomen. Only around seventeen were in common use across the entire Republic. Families reused the same few relentlessly, frequently giving the eldest son the father’s praenomen, so a line could run three or four generations deep on one personal name.

That inverts the modern arrangement entirely. For us the first name carries identity and the surname carries family. For a Roman the personal name carried almost nothing, and everything distinguishing sat in the clan and branch names. Understanding that is the difference between Roman names that feel authentic and Roman names that feel like Latin-flavoured modern ones.

Women were named after the family, not themselves

This is the piece almost all historical fiction quietly abandons, and it is worth knowing what you are abandoning.

A Roman woman typically received a feminine form of her father’s nomen and nothing more. A daughter of the gens Julia was Julia. A daughter of the gens Claudia was Claudia. Sisters shared the name, distinguished where necessary by seniority — the elder and the younger — or by number.

For a novelist this is unusable. A household where four sisters are all called Julia is historically sound and impossible to follow on the page. So nearly everyone cheats, and this tool cheats too: it produces distinct feminine names, because the alternative makes it useless.

The honest way to handle it in fiction is to use the real convention once, visibly, early — let a character be introduced as the younger Julia — and then rely on nicknames and shortened forms thereafter. You get the texture of the real system without the cost.

Cognomina: the nickname that became hereditary

The cognomen is the most interesting of the three because of how it changed over time.

It began as a personal nickname, and the originals were frequently unflattering or merely physical — descriptions of a body, a habit, or an incident. Where modern names are chosen to dignify, the cognomen started as something other people called you.

Over generations these hardened. A nickname that stuck to one notable man passed to his descendants and became the marker of a branch within the gens, at which point it had stopped describing anybody. By the late Republic a cognomen was a piece of family property, and an additional personal one — an agnomen — might be added to distinguish an individual again.

That cycle is worth stealing for invented settings. A name that began as an insult and is now a mark of status carries a whole social history in one word, and it explains why a prestigious family might be called something faintly ridiculous.

Names recorded status, and the system is explicit about it

Roman names encoded where you stood, which makes them unusually useful for a writer. Three mechanisms matter.

Adoption. An adopted man took his adoptive father’s full name, commonly adding a form derived from his birth gens ending in -ianus. The famous case: the man born Gaius Octavius, adopted by Julius Caesar, became Gaius Julius Caesar Octavianus. The name itself records the adoption, permanently and publicly.

Freedmen. A freed slave commonly took the praenomen and nomen of the patron who freed them, keeping their former name as a cognomen. So a man’s full name could announce both that he had been enslaved and who had owned him. There was no way to hide it, and no expectation that one should.

Citizenship. Grants of citizenship worked similarly — new citizens frequently took the nomen of whoever had sponsored the grant, which is why whole provinces can show clusters of the same clan name appearing at once.

The consequence for fiction is that a Roman name is a biography. Introduce a character by their full name in front of someone who can read it, and you have conveyed their origin, their status and their obligations without a line of exposition.

Getting it right in fiction

  • Use the full name rarely. Romans did not address each other with all three names in conversation. Pick which name a given character is called by, and let who uses which signal the relationship.
  • Reuse praenomina on purpose. A father and son sharing a personal name is accurate and immediately legible as a Roman detail. One or two instances carry the texture; more will confuse readers.
  • Let the cognomen do the work. It is the most distinctive of the three and the one readers will remember. If a character needs a memorable name, it should be theirs.
  • Watch the period. Three-part names belong to the Republic and early Empire. Later naming grew far more elaborate and then collapsed into something quite different. A tria nomina in a late-Empire story is an anachronism in the other direction.
  • Check invented names against the record. Latin endings are regular enough that a plausible invention has a real chance of matching an attested person, which matters if your character is unsympathetic.

Why these are invented rather than attested

A generator that returned real names from the historical record would be handing you actual people — and Roman prosopography is well enough documented that many names map to a known individual with a known life.

So this builds names from the structure instead: real praenomen shapes, regular nomen endings, cognomen forms that follow the attested patterns, assembled into people who did not exist. The result is plausible to a reader and not a claim about anyone.

If you need a real Roman, a historical source is the right tool. If you need a Roman-shaped character to put in a story, this is.

How this generator is built

The Roman tradition supplies its own onsets, vowel nuclei, codas and weighted syllable templates, plus the ending sets that make a name read as Latin. Names are assembled from those rather than sampled from a list.

The constraints are tighter here than for the invented traditions, because Latin phonology is regular and a name that breaks it is immediately obvious to anyone who has studied the language. The usual protections also apply: no unsayable consonant runs, no repeated syllables, no badly joined compounds.

The full method is on the how our tools work page.

Other traditions

This is the only historical tradition in the fantasy set; the rest are invented sound systems.

For real modern names across 23 actual naming traditions, with correct name order and gendered surname inflection, the Random Name Generator is the better tool.

Common questions

What is the tria nomina?

The three-part name a Roman citizen man carried: a praenomen (personal name), a nomen (the clan or gens he belonged to) and a cognomen (originally a nickname, later inherited as a family branch name). Gaius Julius Caesar is the standard illustration: Gaius was his personal name, Julius marked him as a member of the gens Julia, and Caesar identified his branch of it.

Why were there so few Roman first names?

Because the praenomen carried very little of the identifying work. Only around seventeen were in common use across the whole Republic, and families reused the same few across generations — often giving the eldest son the father’s praenomen. Identity lived in the nomen and cognomen instead, which is the reverse of how modern Western names work.

This is why Roman names in fiction so often feel repetitive to modern readers. They were repetitive. That was the system working as intended.

How were Roman women named?

Usually with a feminine form of the father’s nomen, and nothing else. A daughter of the gens Julia was Julia; a daughter of the gens Claudia was Claudia. If there were several daughters they were commonly distinguished by order — the elder and the younger, or by number — rather than by separate names.

Modern fiction almost always softens this, because a cast where every sister has the same name is hard to follow. The tool produces distinct feminine names for usability, which is a deliberate departure from strict accuracy.

What happened to names on adoption?

An adopted man took his adoptive father’s full name and commonly added a form derived from his birth gens, ending in -ianus. The famous case is the man born Gaius Octavius, adopted by Julius Caesar, who became Gaius Julius Caesar Octavianus — the Octavianus recording where he came from. It is a neat system: the name itself carries the adoption.

Are these historically accurate names?

They are built on the real structure with regular Latin endings, but they are invented rather than attested. That is deliberate: a generator that handed you real names from the historical record would be giving you actual people, which is a problem for fiction. Use these when you need a plausible Roman who did not exist. If you need a real one, use a historical source.

Can I use these commercially?

Yes. They are assembled from syllables and endings rather than copied from any record. Search anything that becomes a major character, because with Latin endings a plausible invented name has a fair chance of matching someone attested.