The Most Common Last Name in America Depends on Who You Ask

Smith is the most common surname in America, carried by roughly 2.37 million people. That is probably something almost everyone in the United States knows. The more revealing question is what happens when you ask it group by group, because the answer changes almost completely depending on who you ask about.

The U.S. Census Bureau publishes a surname file that reports, for every name held by at least 100 people, the share of bearers who identify as White, Black, Hispanic, Asian or Pacific Islander, American Indian or Alaska Native, or multiracial. Multiply those shares by each name's total count and you get the most common surname within each group. Below is what that produces.

The national top ten, for reference

The most common surname in each group

Now the same question, asked five ways. These are estimated bearers within each group.

Only one group's top name matches the national leader. Among Black Americans, Williams leads, followed by Johnson, Smith, Jones, and Brown, with Jackson sixth at 51.2 percent Black. Among Hispanic Americans the entire top ten is Spanish-origin: Garcia, Rodriguez, Martinez, Hernandez, Lopez, Gonzalez, Perez, Sanchez, Ramirez, Torres. Among Asian and Pacific Islander Americans, Nguyen leads by a wide margin, ahead of Lee, Patel, Kim, and Chen.

The names that belong most to one group

Ranking by raw count favors big names. The sharper measure is concentration: which surnames are most lopsidedly held by a single group. Limiting to names with at least 20,000 bearers, these are the most concentrated.

Each of these has a history attached. Yoder at 96.7 percent White reflects Amish and Mennonite settlement in Pennsylvania and the Midwest; Schwartz (93.7 percent) and Katz (93.3 percent) trace German and Ashkenazi Jewish immigration. Thao and Xiong, both above 97 percent, are Hmong names carried by refugee families who arrived after the Vietnam War. Pierre and Jean, both heavily Black, are largely Haitian American.

The most-cited example is Washington: 172,075 bearers, 83.6 percent Black, the most lopsided of any large classically American surname. Jefferson follows at 71.6 percent. The pattern traces to emancipation, when many formerly enslaved people chose the surnames of founding figures, and to the fact that the names stayed concentrated in the community that adopted them.

Why the American Indian column looks different

One row above is not like the others. The most concentrated American Indian or Alaska Native surname among large names is John, at just 9.6 percent. Why is that you ask?

American Indian and Alaska Native people are roughly 1 percent of the U.S. population, so even a name carried overwhelmingly by Native families rarely clears 20,000 bearers. Drop the 20,000 size floor and the picture changes completely: Begay is 92.8 percent Native with 15,329 bearers, and Yazzie is 93.0 percent with 13,728, both Navajo names. Locklear, at 66.1 percent, is closely associated with the Lumbee of North Carolina. The apparent absence is a function of population size and a threshold, not of distinctiveness.

What the data can and cannot tell you

These are self-reported identities from census responses, aggregated by name, not judgments about any individual. A surname's composition describes a population, and roughly 1/3rd of Smiths are not White, nearly a quarter are Black. Names also blur over time through marriage, migration, and changing identification, which is why the shares shift between census releases.

The Bureau publishes this file precisely because surnames carry demographic history that other statistics miss. It is used in public health and civil-rights research to estimate populations where race is not directly recorded.

You can look up any surname's own page for its full composition and rank history, browse the name rankings, or read our companion piece on the most common last names in America for how the overall top ten has shifted.

Sources

Surname counts and race and ethnicity composition are from the U.S. Census Bureau's surname file, latest release, which covers every surname held by at least 100 people. Group-level counts are estimated by applying each name's reported composition share to its total bearer count. Method notes are on our methodology page.