Back from the 1900s

1. Old1 in 1,000girls or more at its peak, 1880–1935
2. GoneUnder ⅓of that peak at its low, 1950–2005
3. Back1 in 2,000girls or more in 2025, and 3× its low

• The SSA’s national file lists every first name given to five or more babies of one sex in one year, 1880–2025, from applications for Social Security cards. “Per 1,000” divides a name’s babies by every baby of that sex in the file, so names given to fewer than five are left out of both.
• Many people born before 1937 never applied for a card, so the early years count the people who later did. Read the 1880s–1930s as a close sample, not a census.
• The rule: at least 1 in 1,000 babies of its sex at its peak, 1880–1935; a low, 1950–2005, under a third of that peak; and in 2025 at least 1 in 2,000 and three times its low. Peak and low are three-year averages so one odd year can’t qualify a name. William and James never left, so they don’t pass.
• Spellings count separately, as the SSA records them: Sophia and Sofia, Eloise and Elouise are different names here.
• A name’s low year is the year of fewest babies, 1950–2005. Rank is one more than the number of names of the same sex given to more babies that year, so ties share a rank; it matches the SSA’s published top 1,000.
• Quick picks: “Back on top” is the four best-ranked in 2025; “Steepest climbs”, the four in the top 100 that rose most times over from their low; “Next up”, the four fastest risers over five years among names outside the top 100.
• The chart this page started from credits nameyearbook.com; this one is rebuilt from the SSA file. The SSA’s server refuses scripted downloads, so the extract runs in a browser on ssa.gov (scripts/extract-baby-names.browser.js).

Social Security Administration, Popular Baby Names, national data (1880–2025). Data: baby-name-comebacks.csv and baby-name-totals.csv. · Data sources