How these numbers are made

Everything here comes from one public source, and every figure on the site is computed from it in a way you can check. Where a choice could reasonably have gone another way, this page says which way we went.

Where the data comes from

The US Social Security Administration publishes counts of names given to babies born in the United States, from applications for a Social Security number. We use two of its releases: the national series, which starts in 1880, and the state series, which starts in 1910.

The corpus on this site was built on 2026-09-04 from the 2025 release, and holds 117,820 name and sex combinations.

If you download the same files from ssa.gov, those checksums should match. If they don’t, we’re working from a different release than you are.

The five-birth rule, and why gaps are not zeros

The SSA does not publish a name given to fewer than five babies in a year. It is a privacy measure, and it has a consequence worth being clear about: we cannot tell the difference between a name nobody used that year and a name four babies were given.

So we never show a zero where the record is simply silent. On a chart the line breaks. In a table the year is absent, with a note saying why. In words we say “fewer than five”. Filling that gap with a zero would be a claim the source cannot support, and it would make rare names look like they vanished when they only went quiet.

Popularity is a rate, not a rank

Charts use births per million, never rank position. Rank is a poor measure of how common a name is: the gap between #1 and #2 is enormous and the gap between #900 and #901 is a rounding error, but a rank chart draws them the same size.

The denominator is the same-sex cohort. When we say a girls’ name was 7,151 per million, we mean per million girls born that year, not per million babies. That is why our axes say “per million girls born” rather than “per million births”.

Rarity

“1 in 210 girls” is that year’s births of that sex divided by the name’s births, truncated, not rounded. A larger number claims a rarer name, so truncating errs toward the commoner reading. Bridger in 2025 works out at one in 4,459.73 boys and we publish one in 4,459.

Rank

Ranks are competition ranks within a sex: names with identical birth counts share a rank, and the next distinct count resumes at its ordinal position. Two names tied at #10 are both #10, and the next is #12.

Trend

The ten-year figure compares births per million in the latest year with births per million ten years earlier. If a name is missing from either end of that window we show no trend at all, rather than treating the missing year as zero — which would manufacture an infinite rise.

Where a name is used most

We pool the latest three years of state data and compare each state’s rate with the national rate. A figure of 38× means the name is 38 times as common there as it is nationally.

A state needs at least 25 births of that name across those three years before we report it. Below that the rate swings wildly on one or two babies, and it would tell you about sample size rather than about the state. Some names have no qualifying state, and we say so rather than drawing an empty map.

One caveat we owe you: the state files suppress separately from the national file, so they add up to roughly 86% of national births. The state rate and the national rate therefore come from slightly different pools, which nudges these multiples upward a little. We use the national file as the denominator because “how common is this name in the country” is the question the comparison is asking.

Girls, boys, and names given to both

The source records only two sexes, and we don’t pretend otherwise. Where a name is given to both, we say so on the page: at least 25 births of each in the latest year, at a ratio no worse than four to one. In 2025 that describes 343 names.

Meanings

We publish a meaning only when we can cite a source you could go and check. When we can’t, the page says we don’t have a verified meaning — it does not guess, and it does not quietly leave the section blank.

Every meaning we show names its source. Where sources genuinely disagree, we show both readings rather than picking one.

What we don’t know

• Names given to fewer than five babies in a year. They are not in the source at any point.
• Babies without a Social Security number, and names as spelled outside that application.
• Anything about the people. The record is counts of names by year and state, nothing more.
• Nicknames, pronunciation and etymology for most names. Those come from other work, and where we haven’t done it we say so.
• Anywhere outside the United States.

Corrections

If a figure here looks wrong, it may well be. The computation is described above precisely so that it can be checked, and we would rather hear about an error than have it stand.