How to Read a Baby Name Table
Every data table on this site uses the same three measures, and learning what each one can and cannot tell you takes only a few minutes. The count is the number of recorded births given that name in that year, sex, and place. It is the most concrete figure and the only one that is a direct record rather than a comparison. Its weakness is that it cannot be compared across years of different sizes without help: four thousand births in a year with two hundred thousand recorded births is a major name, while the same four thousand in a year with two million recorded births is a minor one.
The rank solves the size problem by converting the count into a finishing position among competitors, but it introduces a distortion of its own: the distance between ranks is not uniform. Near the top of a modern table, a handful of births can separate consecutive ranks, while the gap between rank one and rank two can be thousands of births. A move from rank 90 to rank 80 is therefore not the same kind of event as a move from rank 10 to rank 1, even though both are ten places. The share of births, shown throughout this site as a percentage, is the bridge measure: it is comparable across years of any size, and it preserves the sense of distance that rank discards, which is why a name's peak is described here using all three figures together.
The most common reading mistakes are easy to avoid once named. Do not compare a count from 1880 with a count from 2024 as if the underlying populations matched. Do not treat two spellings as one name, because the SSA does not. Do not assume a name missing from a table was never used: it may simply fall below the five-birth release threshold or below the top hundred cut the table displays. And do not read a single year's movement as a trend; the decade tables exist precisely because naming patterns declare themselves over ten years far more reliably than over one.
A final habit worth adopting is to ask which of the three measures a claim is actually about before accepting it. Statements that a name is back, or finished, or everywhere, usually rest on whichever single measure moved most dramatically, and in a dataset of this size one of the three will almost always have moved dramatically by chance arrangement alone. The tables on this site are laid out so that the check takes seconds: find the name, read across its row, and see whether count, rank, and share tell the same story for the same year. When they agree, the pattern is real in the only sense this data can support. When they disagree, the disagreement itself is the finding, and it usually points to a change in the size of the birth year or in the number of competing names rather than in the standing of the name being examined.
All of the conventions described here are applied uniformly across this site: the name pages, the year tables, the decade tables, and the state pages all draw on the same SSA files, the same release rules, and the same definitions of count, rank, and share. The full source statement, including coverage and limitations, is on the methodology page, and corrections or questions about any figure can be sent through the contact page, where every message is read against the underlying file before a reply is sent.
Disclaimer: Baby Name Data publishes statistics drawn from U.S. Social Security Administration baby name records. Figures are counts of Social Security card applications by year of birth and are shown for general information only. Counts below five in any year, sex, or state are not released by the SSA and are not estimated here. Nothing on this page is advice about what to name a child.
Data: U.S. Social Security Administration baby name records, 1880–2024, retrieved October 2026. Source: ssa.gov/oact/babynames/