About the data

Every figure on this site comes from a source that can be checked. This page sets out what those sources are, what is done to them, and where they stop being reliable.

Where the birth data comes from

United States birth figures come from the Social Security Administration, which publishes counts of names given on social security card applications. Three datasets are used: national totals from 1880 to 2025, state-level counts, and territory counts. Together they run to roughly 8.9 million rows.

The data is a work of the US federal government and is in the public domain. No attribution is required; it is credited here because a figure without a source is not worth much.

What the SSA data is not

This is the single most important thing to understand about every number on this site.

• It is not a census of births. It counts social security card applications. A child born in the US who never had a card applied for does not appear, and neither does anyone born elsewhere.
• Rare names are withheld. Any name given to fewer than five babies in a year is omitted for privacy. A name showing no births in a given year may have been used four times. Absence is not evidence of absence.
• Spelling is flattened. The SSA strips accents, removes spaces and hyphens, and truncates to fifteen characters. "Mary Ann" and "Maryann" collapse into one entry, and diacritics are lost.
• Sex is as recorded on the application, which is not necessarily how a person identifies.
• Figures shift between releases. Earlier years are revised slightly each year. These pages were built from the release covering births to 2025.

How the figures are worked out

Rank

Rank is position within a single sex for a single year, counted the way the SSA counts it: names with equal births share a rank. "Ranked #48 in 2025" means 47 names of that sex were given to more babies that year.

The births chart

Each chart plots recorded births per year across the whole period a name appears, with the peak marked. Where a name has fewer than five years of data the figures are given as a sentence instead — a line drawn through three points suggests a trend the data cannot support.

Trend

The rising, falling or stable label compares the average of the last three years against the average of the three years a decade earlier. A change of more than 20% in either direction sets the direction; anything narrower is stable. A name with no recorded births in the last ten years is marked dormant.

Where in the US

State figures need more care than the national ones, because the under-five rule applies per state per year. A name with 300 births nationally might show 180 across the state files, with the remainder scattered in states that never reached five in a year. Percentages calculated from state data alone would be wrong, and quietly so.

So national totals always come from the national file, state data is used only for relative distribution, and the breakdown is shown at all only where the state records account for at least 70% of a name's national total. On the rest it is left out rather than drawn from a third of the evidence.

Names used for both girls and boys

Where a name is genuinely used for both, it gets a page for each, because one combined rank describes neither. Riley is #48 for girls and #207 for boys, on different curves with different peak years.

The threshold for splitting is that the less common sex accounts for at least 1,000 recorded births and at least 10% of the name's total. Both conditions are needed. An absolute floor alone would split Aaron, which has 4,395 girls against 614,462 boys — real, but 0.7%. A percentage alone would split names with a handful of births either way.

Meanings and origins

Meanings and origins are editorial rather than statistical, and they come from this site's own database, built up over many years. They are being reviewed name by name, and where a meaning has been researched and rewritten the source is recorded against it.

Two things follow from that work:

• Every recorded meaning is shown, not just the first. Many names carry several. Showing one and discarding the rest loses real information, so a name with three meanings displays three.
• Where a name is recorded rather than derived, that is said separately. Appearing in Shakespeare is not an origin. Those names carry a line saying where the name is recorded, and the origin field keeps its own job.

Origins are given as specifically as the evidence allows. Where a name's language or people is known it is named — Yoruba, Igbo, Sanskrit, Dakota — rather than the continent it sits on. Where it is not known, the broader tag stands rather than a guess being made.

Why some names have no page

Not every name in the collection has a page of its own. Names with no US record and nothing distinct to say appear on the A–Z and origin indexes with their meanings, but do not get a page.

This is deliberate. A page carrying a name and eight words of meaning helps nobody, and a site full of them is harder to use, not richer. Names move up to a full page as there is something worth putting on it.

Corrections

If something here is wrong — a meaning, an origin, a figure — it is worth telling us, and it will be checked against the source. Contact details are on the privacy page.

Last updated 7 September 2026.