How many people have the last name Smith? U.S. Census data

2,369,644 people in the United States reported the last name Smith in the 2020 Census. Smith was the most common U.S. surname, followed by Johnson and Williams. These are counts from April 2020, published by the Census Bureau in April 2026; they are not a live count of everyone named Smith today.

The official last-name spreadsheet lets you look up other surnames too. Below, compare the most common last names, find your own surname count, and see why a full-name estimate such as John Smith is less certain.

How many people have the last name Smith in the U.S.?

The 2020 count is 2,369,644, compared with 2,442,977 in 2010. Both are Census counts of a reported surname, not people-search profiles. Name coverage changed between censuses, so the difference does not by itself establish how many people changed their name, moved or died.

How many people have the last name Smith worldwide?

The U.S. Census cannot answer that question. It covers the United States, not all countries where Smith occurs. Do not present its U.S. count as a worldwide total or extrapolate it without comparable country-level data.

How can you find out how many people have your name?

It depends on which part of the name you mean. The 2020 last-name brief is direct: "Data on first and last names are always tabulated separately."

First-name coverage is from the 2020 first-name brief. To look up your last name, open the last-name table and search for it in capitals, without punctuation or spaces: OHARA, not O'Hara. The count is the number of people who reported that name in April 2020.

What are the most common last names in America?

The 2020 top ten, with 2010 counts from the 2010 top 1,000 surnames spreadsheet (XLSX). The last column gives each name's largest Hispanic origin or race groups; "White" and "Black" mean non-Hispanic White alone and Black alone.

The ten names held 4.7 percent of people with a recorded last name in 2020 (2020 brief, Table 4), down from 4.9 percent in 2010. Every count fell, partly because coverage fell from 95.5 percent in 2010 to 90.2 percent.

Census comparison · people counted

The same ten names, two census years

Ranked by 2020 count. Both series start at zero and use the same scale.

20102020

01.25m2.5m

1. Smith
   20102,442,977

   20202,369,644
2. Johnson
   20101,932,812

   20201,858,234
3. Williams
   20101,625,252

   20201,561,395
4. Brown
   20101,437,026

   20201,386,083
5. Jones
   20101,425,470

   20201,383,250
6. Garcia
   20101,166,120

   20201,149,510
7. Miller
   20101,161,437

   20201,129,186
8. Rodriguez
   20101,094,924

   20201,085,838
9. Davis
   20101,116,357

   20201,074,448
10. Martinez
    20101,060,159

    20201,039,848

Coverage changed. A surname was recorded for 95.5% of those counted in 2010 and 90.2% in 2020. Falling counts alone do not establish migration, deaths or name changes.

The chart compares the 2020 counts with 2010 on the same scale. The 2020 last-name brief states the pattern: "In many cases, over 90 percent of people reporting a name identify with just one of the six Hispanic origin or race groups."

How common are the most frequent U.S. surnames?

A few hundred last names cover a quarter of everyone with a recorded last name, and millions are shared by almost no one. Table 1 of the 2020 last-name brief:

The cumulative column of the 2020 last-name table (XLSX) puts the 1,000 most frequent names at 39.7 percent of people.

The 2010 file had the same shape. Its technical documentation, Joshua Comenetz, October 2016, reported about 6.3 million distinct surnames; 11 of them appeared more than a million times; 3.9 million surnames, 61.9 percent of the distinct set, were reported only once. The Bureau also published File B: 162,253 surnames that occurred at least 100 times. Most of the 6.3 million distinct surnames sit below that threshold.

How many people are named John Smith?

A full-name count is estimated by multiplication: a first-name share times a last-name count.

Worked from the 2020 first-name table (XLSX): John was reported 3,125,114 times among 302,029,327 people with a recorded first name, a share of 1.03 percent, or about 1 in 97. Applied to 2,369,644 Smiths, that gives about 24,500 John Smiths.

Why the true count can differ

The multiplication assumes first names and last names are unrelated. They are not, and the error runs both ways. The 2020 tables split every name into six Hispanic origin and race groups, so the arithmetic can run inside each group:

Maria shows how the gap opens. Among everyone with a recorded first name in the 2020 first-name table, 0.55 percent are named Maria; among Hispanic people, 2.4 percent are. Garcia is 91 percent Hispanic, so spreading Maria evenly across all Garcias undercounts. José runs the other way: the first-name brief puts José at 96 percent Hispanic, while Smith is 3.5 percent Hispanic.

The adjusted figure still assumes independence within each group. Every estimate also misses people without a usable name on the form, and the tables describe April 2020: the Bureau's Vintage 2025 estimate is 341,784,857 residents on 1 July 2025.

Figures quoted elsewhere vary more. The CFPB's 2021 advisory opinion, as published in the Federal Register, cites a catalogue putting James Smith at over 30,000 people and quotes a 2016 appeals court putting David Smith at over 125,000, without saying how either figure was made.

What are the limits of Census name counts?

They do not say how many people are named John Smith. The tables do not publish an official count for the full name John Smith.

They also do not describe 2026. The underlying census is 2020. Garcia's rise to sixth place, noted in the 2010 documentation, already showed the list moving with the population.

A 2026 search will not see the same counts. It will see the same shape: a few names shared by millions, a long tail of names shared by almost no one, and a large middle where "uncommon in my circle" is still common in a country.

The edits matter for search. The Bureau deleted punctuation and spaces, so O'Hara and O Hara both become OHARA in the file. Suffixes were stripped, and in 2020 a Ñ became N, so PEÑA counts as PENA. Obvious non-names were removed. A people-search index that keeps the punctuation the person actually uses, or that does not, will miss or merge records for reasons that have nothing to do with identity.

The data "do not in any way identify any specific individuals." That is the Bureau's sentence, and it is the right one. These are aggregates. They tell you about collision risk, not who anyone is.

What a collision looks like in a result

Schematic · an unsupported merge

Three sources do not necessarily describe one person

1. Source AA city
2. Source BA job title
3. Source CA photograph

→

One finished-looking card

A face, a city and a job title, joined only by a shared name.

The unsupported step is the join.

A separate source is useful only when evidence links it to the same candidate. Repeated directory copies do not create independent corroboration.

The failure is a finished-looking result rather than an empty one. Three public pages share a surname: one has a city, one a job title, one a photograph. A product that merges them writes a single card with a face, a timeline and three sources, and asks the reader to believe that two or three people are one.

The Census table tells you how often that setup is available. The simple full-name model above estimates roughly 25,000 John Smiths; that is not a verified count of living people or search results. If they were spread like everyone else, an area holding 1 percent of the population would still have about 250, and the Bureau's 1990 name documentation notes that "families tend to live close to each other".

A rare surname does not rescue the method. Many names reported fewer than 100 times are unique names, variants of more common ones, or transcription errors, the brief says. Two records that share an unusual string can still be different people, or the same person recorded twice, or a relative. Rarity lowers the odds of a collision without proving a match.

How to use the table without misusing it

Use the counts to understand how many candidates a name search might return. They cannot identify a particular person.

1. Estimate the crowd before you search. Multiply the last-name count by the first-name count and divide by 302,029,327, the recorded-first-name total in the 2020 first-name table (XLSX). If both names cluster in the same group, expect more people; if they sit in different groups, fewer.
2. If the estimate runs into the hundreds, do not search the name alone. Add a public anchor you already know: a city, an employer, a school, a handle, a domain, a date range. The common-names guide is the worksheet.
3. If the surname looks rare, still keep candidates separate. A name missing from the table had fewer than 100 reports in 2020 or 2010. It can be a spelling variant of a common name, a typo, or a relative.

Independent evidence means a separate path, not a second copy of the same directory row. Two aggregator pages that reprint one listing are one source. A company page, a personal domain, and a handle that all name the same employer is the beginning of a case. The same rule applies to public records such as court files and deeds.

Methodology

This is a reading of a published official dataset, not an original survey.

• Primary tables: the 2020 Census first-name and last-name tables by race and Hispanic origin, published 14 April 2026 and downloaded 29 October 2026, with briefs C2020BR-13 and C2020BR-14.
• Arithmetic we performed: top-ten sums; independence estimates (last-name count times first-name count, divided by 302,029,327 from the 2020 first-name table); and group-adjusted estimates, the same product within each group divided by that group's total in the first-name table, then added. Estimates are rounded to the nearest hundred.
• What we did not do: estimate middle names; inflate the 2020 counts to 2026; identify any person; or treat DeepSearch results as a sample.
• Historical comparison: the CFPB's 2021 advisory opinion is cited above for the older full-name estimates it quoted, not as a current legal rule.

If you are searching, or being searched

If you are looking for someone, start with the strongest public clue you already have and keep every same-name result in its own row until independent sources agree. How to check a specific claim is How to verify a DeepSearch result using its sources.

If a result mixes you with someone else, that is a correction. Send the result URL, the wrong field, and the source it cites. How to correct or remove information from DeepSearch is the process. Remove my info is the public form.

Frequently asked questions

What is the #1 last name?

Smith, in the United States: first in the 2020 Census and in every earlier list the Bureau compares, back to a 1790 count of the White population. The most common first name in 2020 was Michael. We know of no official worldwide count of names.

What are the 10 most rare names?

The Census tables cannot say. The 2020 last-name brief lists only names reported at least 100 times, and in 2020 over 7 million last names, 98 percent of the distinct names, fell below that line. A name missing from the table is rare in the count, not proven unique.

Is a "how many of me" calculator accurate?

It is as accurate as its inputs. A calculator that multiplies two shares can land close for names spread across groups and miss by four times for pairs that cluster together or apart.