Names That Went Extinct: 2,853 American Names With No Living Bearers Left

In our last post we found that most American names are rare — 67,460 of the 104,819 first names in Social Security Administration records have fewer than 100 estimated living bearers. This is the story we promised at the end of that post: the names at the very bottom of the scale, the ones that did not just fade, but stopped entirely. By our model, 2,853 names — 2.7% of every name ever recorded — now have an estimated zero living bearers. They were real names, given to 30,841 real babies between 1880 and 1930, and then they were never given again.

A name going extinct is not like a word falling out of use. A word survives in books. A name survives only in the people who carry it, and when the last bearer dies, the name leaves the living population for good. The records let us watch that happen, name by name, across 145 years.

What “extinct” means here

We call a name extinct when our survival model estimates zero living bearers for it in 2025 — that is, when the model's estimate rounds to 0. The method is the same one behind the whole site: take every birth recorded for a name in the SSA files from 1880 to 2024, weight each birth year by the probability that someone born in that year is still alive in 2025, and add it up. A name last given in 1930 to five babies, for example, has an expected 0.47 living bearers — five 95-year-olds, each with roughly a 9% chance of still being alive — so it rounds to zero.

Two honest caveats, before the numbers. First, the SSA only publishes a name for a given year if at least five babies received it, so “last appeared” means last year with five or more births — a name could in theory have lingered below that line. Second, these are estimates from a survival model, not death records: we cannot prove that the very last bearer of any single name has died, only that the odds now round to none. What the model can do, reliably, is show the pattern across all 2,853 names at once — and the pattern is strikingly clear.

When the names died: almost all of them in twenty years

Here is when the extinct names were last given to a newborn, by decade:

Not one extinct name was last given after 1930. And 95.3% of them were last given in just two decades, the 1910s and 1920s. That is not a coincidence, and it is not a quirk of the data: 92.8% of the extinct names also peaked in 1910–1929. These were names of a particular moment — the peak years of mass immigration and of a naming culture that had not yet settled into the patterns we know today — and when that moment passed, the names went with it.

The biggest names that died

Most extinct names were tiny: the median one was given to just 6 babies in total, and 1,792 of them — nearly two in three — appear in the records in a single year only. But a few were given to a hundred babies or more, and still died out. The ten largest:

Ragna is the largest extinct name of all: 201 babies over 42 years, peaking with 18 births in 1915, and not one since 1927. No extinct name was ever given to more than 201 babies, and only five were given to more than 100. Extinction, it turns out, is a small-name phenomenon — names that never built a large enough base of bearers to survive a generation that stopped choosing them.

Why they died: three patterns

1. They never took root. The commonest story is the simplest: 2,217 of the 2,853 extinct names — 77.7% — were given over a span of five years or less, and most of those in a single year. These were experiments that lasted one season: a name tried by a handful of families in one town in one year, and never picked up again. Without a second generation of parents choosing the name, there was nobody to carry it forward once the original handful grew old.

2. They arrived with immigrants, and stayed behind. Look at the big names above and a second pattern jumps out: Ragna, Alfhild and Tyyne are Scandinavian and Finnish; Mieczyslaw, Genowefa and Jadwiga are Polish. These names entered the American records with the great immigration waves of the 1900s and 1910s, peaked in 1915–1919, and then faded as the second generation chose American names for their own children. The names did not die in their home countries — they died as American names, because the families who brought them stopped passing them on. It is the same assimilation story, told in birth records instead of census forms.

3. They belonged to a moment. Verdun is the clearest case: a French town that became a byword for the First World War, given to 33 babies in 1918 alone — 25 boys and 8 girls — and never again after 1922. Moment names flare and vanish; the event that made them meaningful passes, and with it the reason to choose them. The extinct list is full of these brief flares, most of them too small to trace to a single event, but Verdun shows the mechanism plainly.

One more thing the list shows: 60.8% of the extinct names are female (1,734 of 2,853). Girls' names have always turned over faster than boys' — parents experiment more with daughters' names, and the fashion cycle runs quicker — and the extinction record is that churn, preserved.

Next to go: the names with one foot in the grave

Extinction does not happen all at once; there is a queue. Of the 5,965 names last given before 1950, 2,853 are already at zero — and another 2,746 sit at an estimated 1 to 9 living bearers, with 366 more at 10 to 99. These are the names whose last bearers are now in their nineties, and whose lines will run out in the coming years. A few of the largest:

Hildur was given to 718 babies — more than three times as many as Ragna, the largest fully extinct name — yet only about three bearers are left, all of them born before 1934. Size alone does not save a name; what matters is whether anyone is still choosing it. None of these names has been given to a baby in living memory, and when their last bearers go, they will join the 2,853.

There is a counterweight, and it is worth remembering from the last post: extinction is only half the story. New rare names are invented every year — 7,573 rare names last appeared in the birth records in 2024 — so the total stock of names keeps growing even as old ones die. A century from now, some of today's five-baby names will be the next Michaels, and most will be the next Bacilias. The churn never stops; only the individual names do.

How we count “living bearers”

For each name, we take every birth recorded in the SSA files from 1880 to 2024 and weight it by the probability that someone born in that year is still alive in 2025, using a cohort survival curve interpolated from SSA life-table patterns. Add up the weighted births and you get the estimated living bearers; round to the nearest whole person, and 2,853 names land on zero. The full method — formulas, the survival curve, and the known limits — is on our methodology page.

The figures are estimates, not a census — but they are computed the same way for every name, so the comparisons between names are solid, and the headline count reproduces exactly from the raw SSA files.

Check your own name

If this has you wondering where your own name sits — somewhere between Michael's 3.6 million and Ragna's zero — you can find out in seconds. Run it through the free checker on our homepage, or use the name rarity checker to see its rarity score and how many people in a million share it. Curious about a full name? The name combination checker estimates how many people share your exact first-and-last-name pair. And if you want to see how a name rose or fell over time, the baby name popularity by year tool charts all 145 years of it.

Most names are rare names. A few thousand are gone entirely — and now you can put a number on exactly how rare yours is, before it joins them.