About the Baby Name Finder
Most “baby name generators” are a list of forty names with a one-word meaning and no data behind them. This is the other thing: the Social Security Administration’s national dataset — every first name given to five or more babies in a year, 1880 to the latest year published — turned into a browser. Filter by how rare a name is today, whether it is rising or falling right now, the decade it peaked, its first letters or its length, and every result carries its real current rank, last year’s count, its peak year and its five-year trend. Click any name for its whole history.
The subscriber layer is where the data earns its keep. Sibling-style matching finds names whose popularity curves have the same shape and era as a sibling’s, ranked by cosine similarity — a computed definition of “goes together”. The comeback list holds only names whose curve actually turned: a pre-1950 peak, a real lull, three-plus years of growth. And the shortlist is shareable, so a partner can rank it and send it back. What it does not have, it says: no regional data, no meanings, no origins — just the numbers, which are the useful part.
Heart a name to add it. Rank with the arrows; the order is saved in this browser. Share the list and your partner can rank it their way and send a link back.
Names whose popularity curve has the same shape and era as a sibling’s — same rarity band today, peak within twelve years, ranked by how closely the decade-by-decade curves match.
Grandparent-era names whose curve has genuinely turned: peaked in 1950 or earlier, fell below 5% of that peak, and have now risen three or more years running. Computed from the counts, not asserted.
How to Use the Baby Name Finder
Pick girls or boys, then narrow: a rarity band for how common the name is today, a trend for whether it is on the way up or down, the decade it peaked, a first letter or two, a length, or the year you were born to see what peaked with you. Every result shows its rank and count in the latest year, its peak year and its trend; the name itself links to its full history since 1880. Sort by current popularity, by the longest rising streak, by all-time total, by newest peak, or alphabetically.
Rare Is Not the Same as Unheard-Of
The rarity bands are rank ranges in the latest year, and the most useful one is the least dramatic. A name at rank 800 is given to perhaps 350 babies a year nationally — about seven per state — which is exactly the sweet spot most parents mean by “not too common”: instantly recognisable, spelled without help, and almost never duplicated in a classroom. The “Rare” band goes further, down to names given to a handful of babies, and it is where the finder differs from a generator: every name in it was chosen by real parents this decade.
Trends Are Measured, Not Announced
A name is rising if its share of births grew at least 15% over five years, falling if it shrank at least 13%. The streak counts consecutive years of growth, and the “rising fastest” sort uses it, because a five-year jump can be one celebrity baby while a nine-year streak is a genuine movement. Rising names skew toward the uncommon bands — by the time a name reaches the top ten its growth has usually slowed — so a parent hunting for “the next big thing” should filter Uncommon plus Rising and sort by streak.
The Hundred-Year Rule, Tested
The folk rule says names come back after about a century, when the people who wore them are grandparents rather than parents. The data broadly agrees: Emma peaked in the 1880s and returned to #1 in 2008; Henry, Theodore, Eleanor and Hazel all sat out the middle of the twentieth century and are top-fifty names again. The subscriber comeback list applies a strict version of the rule — an early peak, a real lull, and three or more years of measured growth — and currently admits thirty-five names, which is the point: a shortlist of what is actually turning, not everything that sounds old-fashioned.
Sibling Sets That Match by Data
Parents want siblings’ names to “go together”, and the finder gives that a definition: names with the same rarity today, a peak within twelve years of each other, and the most similar decade-by-decade curve. Ask for a brother for Eleanor and the numbers return Jack, Leon, Calvin, Arthur and George — the same 1910s peak, the same 1980s lull, the same 2010s revival. It will never suggest a name because it “feels” right; it suggests the ones whose history rhymes.
What Is Deliberately Missing
Meanings and origins, because the SSA does not record them and the usual sources are unreliable; a regional filter, because the state-level files are not bundled; and any name given to fewer than five babies in a year, because the SSA withholds them. Spellings are separate rows — search “Sof” and both Sofia and Sophia appear with their own numbers — and the finder leaves the merging to you.
Related tools: Your Name Through Time for any single name’s full history, Birthday Twins for who shares the due date, and the Ovulation Calculator for the earlier part of the plan. Browse every Fun & Novelty tool for more.
Source: Social Security Administration national baby-names data, 1880 through the latest year published (names given to five or more babies of one sex per year), refreshed each May. Bands use latest-year competition ranks; trends use five-year change in share of births; comeback and sibling matches are computed from the year-by-year counts. National data only; spellings are separate.
Frequently Asked Questions
What do the rarity bands actually mean?
They are rank ranges in the latest year the SSA has published. Top 10 and Top 100 are what they say. "Top 500" covers ranks 101–500, roughly names given to several hundred babies a year. "Uncommon" is ranks 501–2,000 — recognisable, but a child will rarely share it with a classmate. "Rare" is everything below rank 2,000 that still reached the five-baby reporting floor: real names in real use, given to between five and a hundred or so babies nationally. Nothing is invented to fill a band.
How is "rising" decided?
By the change in a name's share of births over the last five years — up at least 15% is rising, down at least 13% is falling, anything between is steady. Share, not raw count, so a national dip in births does not make every name look like it is fading. The "rising fastest" sort uses the streak: how many consecutive years the share has grown, which separates a genuine run from a single-year spike after a celebrity baby.
What is a "sibling-style" match?
Names whose popularity curves have the same shape and era as the sibling's — computed, not curated. The tool takes the sibling's decade-by-decade share of births, finds names in the same rarity band today that peaked within twelve years, and ranks them by cosine similarity of the two curves. Eleanor, for instance, brings back Jack, Leon, Calvin, Arthur and George for a boy — names with the same 1910s peak, 1980s lull and 2010s revival. It is a data definition of "goes together", and it says so.
How does the comeback list work?
A name is flagged as a comeback only if the curve shows one: it peaked in 1950 or earlier with at least 500 births, fell below 5% of that peak share during 1985–2005, and has now risen for three or more consecutive years to at least 25 babies. Thirty-five names pass — June, Goldie, Opal, Rosie, Florence, Louise, Cleo, Clara, Mabel among them. Nothing is on the list because it sounds vintage; everything is on it because the numbers turned.
Are these US names only?
Yes: the source is the Social Security Administration, so it reflects babies registered in the United States, and spellings are counted separately (Sofia and Sophia are two rows). It does not include names given to fewer than five babies of one sex in a year, and it is national only — the SSA's state files are not bundled, so there is no regional filter here rather than an invented one.
