Name Popularity Explorer

Why share matters more than counts

The measure switch above changes the answer, and it is worth understanding which one you want.

Raw counts tell you how many babies got a name in a year. That sounds like the obvious measure and it quietly misleads, because the number of births in the United States has moved enormously. There were far more babies born in the late 1950s than in the 1890s, so mid-century names appear dominant partly because there were simply more children.

Share fixes that. It expresses the count as a percentage of all babies of that sex born the same year, which is the only way to compare a name in 1900 with a name in 2020 honestly. Use share for comparing eras, and counts only when you want the absolute number.

The gap can be dramatic. A name may show a higher raw count in 1960 than in 1890 while having been several times more common, proportionally, in 1890.

What the shapes tell you

Charted over a century, names fall into a small number of recognisable shapes, and the shape tells you more than the peak does.

The single spike

A sharp rise and an equally sharp fall, usually over fifteen or twenty years. This is a name attached to a specific moment: a film, a song, a public figure. Names with this shape date their bearer precisely, which is either exactly what you want in fiction or exactly what you want to avoid for a child.

The long plateau

Decades of steady use with no dramatic peak. These are the genuinely classic names, and they are rarer than people assume. A plateau means the name never became a fashion, so it never became dated.

The slow collapse

A high early peak followed by a gentle century-long decline. Most Victorian and Edwardian favourites have this shape. They read as old rather than as dated, which is a different thing and often more usable.

The return

A peak, a long absence, and a fresh rise. There is a rough rhythm to this, often described as a hundred-year cycle: a name feels dated while the generation that carried it is still around, and fresh again once it is not. The discovery panel above can find these directly.

The handover

Some names cross between sexes, and the chart shows the handover happening. Add the same name for both girls and boys and you can see one line fall as the other rises. The crossing is usually faster than people expect, and it tends to run in one direction.

How to actually use this

Naming a child

The useful question is not whether a name is popular but whether it is rising. A name at 0.3% and climbing will be far more common in your child’s classroom than one at 0.5% and falling, because the classroom is filled from a single birth year.

Chart your shortlist together. Names that look similar in a book often have completely different trajectories, and five minutes here tells you something a popularity ranking cannot.

Writing fiction

This is the fastest way to check whether a character’s name fits their age. Work out the birth year, chart the name, and see whether it was in use. A seventy-year-old with a name that did not appear until 1995 is the kind of error readers notice without being able to say why.

The discovery panel is useful in the other direction too: filter for names that peaked in the decade your character was born and you get a list of period-accurate options in one go.

Genealogy and research

A name’s popularity curve helps date an undated record. If a name was barely used before 1920, an ancestor carrying it was probably not born in 1890. It is weak evidence on its own and useful alongside everything else.

Reading the data honestly

Three limits worth holding in mind, because they change what conclusions the chart supports.

Rare names are missing. SSA excludes any name given to fewer than five babies in a year. The dataset is authoritative about what was common and silent about what was unique. A flat line at zero means "fewer than five", not "none".

Spelling variants are separate names. Every alternative spelling counts as its own entry. A name that looks moderately popular may be considerably more common once you add its variants together, which the chart will not do for you. Add them as separate lines and read them together.

Early years are less reliable. The dataset begins in 1880, but Social Security numbers were not issued until 1936. Early years are reconstructed from people who applied later in life, so they under-represent anyone who died young or never applied. Treat pre-1937 figures as indicative rather than exact.

Related tools

The Random Name Generator uses the same dataset to produce names weighted by what was actually common in a decade you choose. The Baby Name Combiner blends family names into something new.

Common questions

Where does this data come from?

The US Social Security Administration, from card applications covering every birth year from 1880 to 2025. It is a public-domain dataset of roughly 375 million births. We process it into compact files at build time; the chart itself runs in your browser.

Why can I not find a name I know exists?

Two possible reasons. SSA excludes any name given to fewer than five babies in a year, to protect privacy, so genuinely rare names never appear. And we chart the 4,000 most-used names rather than all 106,000, to keep the download small. Common names are all here; unusual ones may not be.

What is the difference between share and number of babies?

Number of babies is the raw count. Share is that count as a percentage of all babies of that sex born in the same year, which is almost always the more honest view. Far more babies were born in 1957 than in 1890, so raw counts make mid-century names look artificially dominant. Switch to share to compare eras fairly.

What do the patterns in the discovery panel mean?

They are computed from each name’s own history. All but gone means recent use is under a twentieth of the name’s peak. Came back means the peak was over 45 years ago but the name is being used again at a meaningful rate. Rising means the peak is recent and current use is close to it. The exact thresholds are on our how our tools work page.

Does this cover countries other than the United States?

Not yet. American records are unusually complete and unusually open, which is why almost every tool of this kind uses them. We would like to add UK data next, since the Office for National Statistics publishes something comparable, and the chart is already built to take a second country.

Why does a name appear for both girls and boys?

Because it was genuinely given to both, and SSA records them separately. Names that cross over are common and the crossings are often the interesting part: several names now read as strongly feminine were majority masculine within living memory. Add both versions to the chart and you can watch the handover happen.