Race and Ethnicity Survey Questions

The 2024 US federal standard: one combined question

This is the current shape for anything that needs to line up with US federal data. The 2024 revision to Statistical Policy Directive No. 15 merged race and ethnicity into a single select-all question and added Middle Eastern or North African as its own category, where the 1997 standard had counted those respondents as White. If you are designing a new survey today and you work with US data, start here.

• What is your race and/or ethnicity? Select all that apply.

  Select all that apply
  American Indian or Alaska NativeAsianBlack or African AmericanHispanic or LatinoMiddle Eastern or North AfricanNative Hawaiian or Pacific IslanderWhitePrefer to self-describePrefer not to say

  The seven categories here are the minimum reporting categories in the 2024 standard, in its order. "Race and/or ethnicity" is the recommended stem — it tested better than asking about either word alone.

• Which of these describes your Hispanic or Latino origin? Select all that apply.

  Select all that apply
  MexicanPuerto RicanSalvadoranCubanDominicanGuatemalanAnother originPrefer not to say

  The 2024 standard also asks for detail below the top-level categories. A follow-up like this is how you collect it without making the first question enormous.

• Which of these describes your Asian origin? Select all that apply.

  Select all that apply
  ChineseAsian IndianFilipinoVietnameseKoreanJapaneseAnother originPrefer not to say

• Which of these describes your Middle Eastern or North African origin? Select all that apply.

  Select all that apply
  LebaneseIranianEgyptianSyrianMoroccanIsraeliAnother originPrefer not to say

• How would you describe your race, ethnicity, or origin in your own words?

  Short text

  Always offer this. Fixed categories flatten national origin, tribal affiliation, and mixed heritage, and the free-text answers are where you find out which of your options is doing the flattening.

The older two-question format

The 1997 standard asked Hispanic origin first as a separate yes/no, then race from five categories. You will still meet it constantly, because most historical datasets use it and agencies have until 2029 to finish moving off it. Use it only when you must match existing data — and if you do, say so in your methodology, because the two shapes do not produce comparable numbers.

• Are you of Hispanic, Latino, or Spanish origin?

  Single choice
  YesNoPrefer not to say

  The 1997 ethnicity question, asked before the race question. Under the 2024 standard this is folded into the combined question instead.

• What is your race? Select all that apply.

  Select all that apply
  American Indian or Alaska NativeAsianBlack or African AmericanNative Hawaiian or Other Pacific IslanderWhitePrefer not to say

  The 1997 five categories, which is where the "Other Pacific Islander" wording comes from. Note that Middle Eastern and North African respondents were counted as White under this standard — a known source of undercounting.

• Which category do you identify with most?

  Short text

  Included as a warning, not a recommendation. Forcing multiracial respondents to pick a single primary category is the most common way surveys produce unusable race data. If you need one value per person for a chart, derive it in analysis and document how — do not push the decision onto the respondent.

• Do the options above describe you?

  Single choice
  YesPartlyNoPrefer not to say

  A cheap validity check when you are stuck fielding an older category set.

International and non-US surveys

US federal categories are for US data and do not travel. The UK asks a single ethnic-group question with an entirely different structure; many countries do not collect ethnicity at all, and in some it is legally restricted. If your respondents are international, ask country first and branch — a US category list shown to everyone produces both offence and noise.

• Which country do you live in?

  Short text

  Ask this before any ethnicity question on an international survey. It decides which category set is even appropriate.

• What is your ethnic group?

  Single choice
  WhiteMixed or multiple ethnic groupsAsian or Asian BritishBlack, African, Caribbean or Black BritishOther ethnic groupPrefer not to say

  The high-level shape used in UK data collection, with detailed sub-options normally offered under each heading. Not interchangeable with the US categories.

• What is your nationality?

  Short text

  Nationality is citizenship, not ethnicity, and the two are routinely confused in survey design. Ask whichever one your analysis actually needs, and do not treat an answer to one as an answer to the other.

• In which country were you born?

  Short text

• What language do you mainly speak at home?

  Short text

  Often the more actionable question. It tells you whether your survey should have been translated, which nationality does not.

Reducing refusal and checking your options work

Refusal on these questions is information, not noise: if a fifth of your respondents skip the ethnicity question, that fact belongs in your write-up. Say why you are asking and who sees the answers before the question rather than after, keep it optional, and give people a way to tell you the options were wrong.

• Did the options include a category that fits you?

  Single choice
  YesPartlyNoPrefer not to say

  The fastest way to find a broken option set. Anything above a few percent of "no" means the list needs work before the next wave.

• How comfortable were you answering these questions?

  Single choice
  1 - Very uncomfortable23 - Neutral45 - Very comfortable

• Would you rather skip questions about your background?

  Single choice
  YesNo

  Offering the exit explicitly tends to raise completion on the questions that follow, because it removes the sense of being cornered.

• May we include your answers in aggregate reporting?

  Single choice
  YesNo

• Is there anything about how we asked these questions you would change?

  Long text

• What do you consider most important to how you identify?

  Short text

  Optional, and genuinely revealing in qualitative research — people frequently name something no category list would have captured.

Try it live

The same questions, asked one at a time

This is a working form built from the questions above — answer it the way a respondent would, then take it into your own account with one click.

See how conversational forms ask these one at a time →

Frequently asked questions

Should race and ethnicity be one question or two?

One, if you are designing a new survey for US respondents. The 2024 federal standard uses a single combined select-all question, which testing found produces more accurate answers than asking Hispanic origin separately and then race. Use the older two-question format only when you need to match a historical dataset that was collected that way.

Should respondents be able to select more than one category?

Yes, always. A single-choice race question forces multiracial respondents to misreport themselves, and the error concentrates in exactly the groups most surveys are trying to measure. If your analysis needs one value per person, derive it afterward and document the rule you used.

Can I use the US categories for an international survey?

No. The categories are built for US data and do not describe how people identify elsewhere; the UK uses a different structure entirely, and some countries restrict collecting ethnicity data at all. Ask country first, then branch to a category set appropriate to that country, or ask an open free-text question instead.