In the 2023 U.S. first-divorce estimates, rates ranged from 6.4 to 21.3 per 1,000 men and 7.9 to 25.7 per 1,000 women in first marriages, ages 18 and older. The National Center for Family & Marriage Research reported six race and ethnicity groups. These annual rates cannot tell you what share of marriages will eventually end in divorce.
Read what marriage-duration figures count→
On this page
Who got counted in the 2023 rates
The 2023 profile uses the American Community Survey, a U.S. population survey. It reports first-divorce rates for men and women ages 18 and older. The first-marriage measure groups adults by their own race and ethnicity. The figures do not describe the racial or ethnic pairing of two spouses.
The earlier profile’s definition is first marriages ending in divorce per 1,000 first marriages. Keep the first-marriage scope beside these figures. A rate that includes remarriages answers a broader question. A lifetime percentage needs evidence that follows marriages over time.
U.S. first-divorce rates by race and ethnicity, 2023
The table may be wider than the screen. Focus it and use the arrow keys to reach every column.
| Source group | Men, per 1,000 | Women, per 1,000 |
|---|---|---|
| White, non-Hispanic | 10.2 | 11.5 |
| Black, non-Hispanic | 21.3 | 25.7 |
| Asian, non-Hispanic | 6.4 | 7.9 |
| Hispanic | 13.3 | 15.7 |
| Other, non-Hispanic | 14.4 | 18.1 |
| Two or more races, non-Hispanic | 15.3 | 23.0 |
“Denominator” means the full group used to calculate a percentage or rate.
How to read it: Source: NCFMR Family Profile FP-25-02, Figure 3; 2023 one-year American Community Survey, U.S. adults ages 18 and older. First-divorce rates per 1,000 men or women in first marriages. The source combines Asian with Native Hawaiian and Other Pacific Islander identities. Other includes American Indian and Alaska Native identities because of small samples. The chart gives no confidence intervals. These are published estimates, not calculations from individual records.
Download the table (CSV) ↓Read a rate per 1,000
For the 2023 non-Hispanic Black women’s estimate, the published rate is 25.7 per 1,000. By Research Desk arithmetic, 25.7 divided by 1,000 and multiplied by 100 equals 2.57%. That conversion retains the annual first-marriage measure. It does not turn the estimate into a lifetime chance of divorce.
Use the same unit when you compare rows. An annual first-divorce rate counts a different group from the share of adults currently divorced or the lifetime share of marriages ending. If a headline omits the denominator, find the original table before comparing its number with another.
Keep age and survey year in the comparison
A separate profile published in 2025 gives age-specific first-divorce rates using the 2019 five-year survey file. The author chose that file to reduce sampling uncertainty and account for a change in the race question in 2020. Its publication year does not make those observations 2025 data.
That older analysis reports variation within race and ethnicity groups across ages. An overall group rate leaves those ages combined. The six-group chart does not isolate a causal effect of race, and its displayed values have no attached confidence intervals. Avoid using small differences as proof that one group has a different underlying rate.
Choose the statistic that answers your question
For an annual first-divorce comparison, keep the source year and group definition with the table. For a question about how long marriages last, read the marriage duration and survival figures. For a geographic comparison, use divorce rates by state.
The U.S. marriage measures explain other ways to count marriage and divorce. Choose a source by its population and time period before you use its headline number. These group estimates cannot tell you whether a particular relationship will end or explain a partner’s motives.
