Compare your surname across censuses
The 2010–2020 comparison puts both snapshots beside each other. Each year has its own recorded-surname denominator, including unlisted names. The bars compare frequency per 100,000 records, so a larger total population does not automatically make a name look more common. Save both rows as a CSV, or download the entire 2020 table.
The Census Bureau released the 2020 names tables in April 2026. Their counts include privacy adjustments; they are not exact counts of people alive today. The release’s stated rule requires at least 100 records in both censuses before adjustment, which can leave some published counts below 100. A missing entry does not establish that a name disappeared. The comparison method explains the coverage differences and the source’s one unmatched spelling.
Your last name, measured against a real population
The U.S. Census Bureau published 162,253 surnames appearing at least 100 times in its 2010 surname tabulation. This lookup gives the recorded count, official rank and approximate “1 in” frequency for a spelling. It is a historical snapshot: it does not add births, subtract deaths or estimate migration since 2010.
The denominator is 294,979,229 people with recorded surnames, including the combined count for names below the published threshold. That is smaller than the entire 2010 Census population because not every record supplied a usable surname. For example, Smith appears 2,442,977 times, or about one in 121 surname records. See the surname method and original documentation for the exact calculation.
Browse and compare last names
The surname explorer covers the full published table. Choose a letter, enter a prefix, or limit the recorded count to find less common spellings. “Lowest published counts” starts with names recorded 100 times. These are the smallest released counts, not a list of the rarest names in existence. Many spellings are hidden below that threshold.
The rank printed beside each name is its official Census rank across the complete table. Filtering to a letter or count range does not make the first result rank #1. Equal counts are ordered alphabetically while keeping their source ranks.
Compare two surnames to see their counts, ranks and frequency per 10,000 records. Both use the same denominator. If either spelling is missing, the comparison stays unknown. You can save a comparison image with the historical source attached.
For character or story names, the last name generator samples matching spellings at random. Filter by starting and ending letters, length or frequency, then copy or save a shortlist with the Census counts attached.
Why familiar surnames appear so often
The surname atlas groups records into five non-overlapping rank bands and the unpublished remainder. About 41.0% of the recorded population falls within the top 1,000 surname ranks. At the other end, 29,312,001 records belong to names below the publication threshold. The familiar names at the top therefore describe only part of the picture.
The most common last names in America are ranked there with counts, and the 1,279 rare last names recorded exactly 100 times are listed in full. Each square in the infographic represents approximately 1% of the recorded surnames; exact counts and percentages accompany the chart. The downloadable version keeps the year, denominator and source. Frequency alone cannot explain where a surname originated, which places it is common in, or whether two people are related.
How the lookup treats spelling
Capitalization does not change the answer. The Census table spells every surname with the letters A–Z only: its editors deleted punctuation and spaces, so O’Brien, O BRIEN and OBRIEN reach the same entry. Accents are not part of the table either, so Muñoz is listed as MUNOZ and García as GARCIA; the result tells you when that happens. The table also keeps at most 15 letters, so a longer surname is compared with the entry that holds its first 15 letters. Your result still shows the surname as you typed it. Different letters are never substituted, and names in other scripts are not transliterated.
The Census applied editing rules before publication, including removing some suffixes and combining punctuation variants. A spelling that is not among the 162,253 published names shows as “Too few to list”, with the 100-person rule beside it, and you can still put it on a card. It remains an uncertainty, not a count the tool can quantify. It would be misleading to turn every absent name into “one of a kind.” The released records do not provide that evidence.
A first name answers a different question
The first-name calculator compares baby-name records within one birth year. This surname calculator covers recorded residents in a single Census. The two populations differ, so the numbers should not be multiplied to estimate your full name’s rarity. The combined card can display both, with each source kept beside its own result. You can also add birthday rarity for a third independently labeled fact.
For a separate Census-based first-name lookup, use the 2020 first-name table, which retains its own release and denominator.
Your typed surname stays in the browser. The site downloads the public Census table, performs the lookup locally and exports an image only when you choose to save or share it.
Curious about the language you use at home? The language rarity tool compares published U.S. home-language estimates and makes a separate share card, with its own source and period.
Good questions
Can I compare my surname in the 2010 and 2020 censuses?
Yes. The census-year comparison shows each release’s count, rank and frequency per 100,000 recorded surnames. The 2020 counts are privacy-adjusted, and coverage and editing changed between censuses. The difference describes published counts, not exact population growth. Missing names stay unknown.
How common is my last name?
Type it into the lookup above for its count, rank and share of the 294,979,229 people with a recorded surname in the 2010 Census. Smith, the most common, belongs to 2,442,977 people, about 1 in 121. A surname shared by 1,000 people is about 1 in 295,000, and the rarest published surnames belong to exactly 100 people.
How many people have my last name?
The 2010 Census counted the people with each of 162,253 surnames shared by at least 100 people, and the lookup above shows yours. Surnames with fewer than 100 people are not published, so no count can go lower than 100.
Which country does the surname result cover?
The United States, using surname records from the 2010 Census. It is not a worldwide estimate or a current population count.
What if my surname is not listed?
The public table lists surnames occurring at least 100 times. An absent spelling may have fewer records, be edited differently, or be missing from the source. The calculator cannot assign it a reliable count.
Does a shared surname mean we are related?
No. Surname frequency alone does not establish kinship, ancestry, ethnicity or the origin of an individual’s name.
Can I browse rare surnames or compare two last names?
Yes. Browse all 162,253 published spellings by first letter, prefix or count, or compare two surnames using the same 2010 Census population. The smallest published count is 100; names below that threshold cannot be ranked individually here.