Divorce Statistics by Gender, Age, and Race in the US

The claim that “half of all marriages end in divorce” has been repeated so often it has taken on the status of settled fact — but the reality of American divorce statistics is considerably more nuanced. The overall rate has been declining for decades, varies dramatically across demographic groups, and means different things depending on which measurement methodology you use. Understanding the actual data requires knowing how divorce is counted, which populations are most and least at risk, and how those patterns have shifted over time.

This article draws on data from the CDC’s National Center for Health Statistics, the US Census Bureau’s American Community Survey (ACS), and the National Survey of Family Growth (NSFG) to give an accurate, current picture of divorce rates broken down by gender, age, and race and ethnicity.

Q: Is it true that 50% of marriages in the US end in divorce? A: This figure has been outdated for years. The crude divorce rate — divorces per 1,000 people — peaked around 1981 at approximately 5.3 per 1,000 and has declined steadily since, reaching approximately 2.4 per 1,000 in the most recent CDC data. Lifetime divorce risk for recent marriage cohorts is estimated at around 40–45%, not 50%, and is declining further as marriage itself becomes more selective — people who marry today tend to be older, more educated, and more economically stable than those who married in the 1970s and 1980s when divorce rates peaked.

1. How Divorce Data Is Collected in the United States

Divorce statistics in the US come from two primary sources, and they measure slightly different things. The CDC’s National Center for Health Statistics collects divorce counts through the Vital Statistics Cooperative Program, in which states report divorce certificates to the federal government. However, several large states — including California, Georgia, Indiana, and Louisiana — have not consistently reported divorce data to the CDC for extended periods, creating gaps in the national dataset that require estimation.

The Census Bureau’s American Community Survey takes a different approach, asking individuals about their marital history as part of a large annual household survey. The ACS produces estimates of the number of currently divorced people (those who have not remarried), the annual probability of divorce, and lifetime divorce risk for different demographic groups. Because the ACS asks individuals directly, it captures divorces that may not be recorded in state-level vital statistics.

When reading divorce statistics, it is worth paying attention to which metric is being used:

• Crude divorce rate: divorces per 1,000 total population (includes children, the unmarried, etc.)
• Refined divorce rate: divorces per 1,000 married women (a more accurate measure of risk)
• Cumulative lifetime divorce probability: estimated percentage of marriages that will eventually end in divorce

The crude rate is often cited because it is simple, but the refined rate and cumulative probability are more meaningful for understanding actual divorce risk.

2. Overall US Divorce Rate: Where It Stands Today

The most recent CDC national divorce data — from the National Vital Statistics Reports — shows a crude divorce rate of approximately 2.4 per 1,000 population and a refined rate of approximately 14–15 per 1,000 married women. Both figures represent a substantial decline from peak levels and place the US divorce rate below its historical highs, though still above the rates seen in most Western European nations.

The long-term trend is clearly downward. The crude divorce rate has fallen from 5.3 (1981 peak) to approximately half that level today. Researchers attribute this decline primarily to:

• Later age at first marriage: The median age at first marriage has risen from approximately 22 for men and 20 for women in 1980 to approximately 30 for men and 28 for women today — and later marriage age is one of the strongest predictors of lower divorce risk
• Increased educational attainment: College-educated people marry later and divorce at substantially lower rates
• Selection effects: As marriage rates themselves have declined, those who do marry tend to have stronger demographic predictors of stability
• Economic factors: The financial cost of divorce is a deterrent, and extended cohabitation before marriage has sorted out incompatible couples before legal marriage

The “50% of marriages end in divorce” figure was a projection based on the high divorce rates of the 1970s and early 1980s applied forward — a projection that has not materialised because the social conditions that produced that peak did not persist.

3. Divorce Statistics by Gender

Divorce data by gender reveals a consistent and important asymmetry: women initiate approximately 69–75% of all divorces in the United States, according to research published in the American Sociological Review and consistent with findings from the National Survey of Family Growth. This figure has been replicated across multiple study designs and time periods.

However, initiation rate does not translate directly to equal impact. Post-divorce economic outcomes differ substantially by gender:

• Women’s household income drops an average of 20–30% following divorce, according to multiple studies including research from Stanford’s Center on Poverty and Inequality
• Men’s household income drops a smaller amount on average, though outcomes vary significantly based on whether they pay child support and alimony
• Custodial responsibility: Women still bear primary physical custody in the majority of cases following divorce — approximately 65% of custodial parents are mothers — though shared physical custody arrangements have become significantly more common since the 1990s

The higher female initiation rate is typically explained by researchers as reflecting the fact that women — particularly in heterosexual marriages — bear a disproportionate share of the dissatisfaction that builds over time in troubled marriages. Research by sociologist Michael Rosenfeld found that women’s higher initiation rate applies primarily to marriages but not to non-marital relationship breakups, suggesting it reflects something specific to the institution of marriage as currently structured rather than a general pattern of female relationship dissatisfaction.

The gender asymmetry in divorce initiation — approximately 70% female — is one of the most consistently replicated findings in the sociology of marriage, and it fundamentally challenges the popular narrative that divorce is something that happens to women rather than something they actively choose.

4. Divorce Statistics by Age Group

Age at marriage is one of the strongest predictors of divorce risk, and the relationship is not linear in the way many people assume. The data on divorce rates by age group reveals several patterns worth understanding:

Teen marriages (under 20): Have the highest divorce risk of any age group. Marriages where at least one partner is under 20 at the time of marriage divorce at roughly double the rate of marriages where both partners are in their late 20s or older. In the NSFG data, approximately 48% of marriages where the woman married before age 18 had dissolved within 10 years.

Early 20s marriages (20–24): Also carry significantly elevated divorce risk compared to later-married couples. The NSFG estimates approximately 35–40% of marriages in this age group dissolve within 10 years.

Mid-to-late 20s marriages (25–29): Substantially lower divorce risk — approximately 25–30% 10-year dissolution rate. This is close to the median range for stable marriage outcomes.

Age 30+ marriages: The lowest divorce risk by 10-year measurement, though this partly reflects selection effects — people who marry at 30 or later have typically had more relationship experience, are more financially stable, and are making a more deliberate choice.

The concept of “grey divorce” — divorce among couples aged 50 and older — has increased significantly as a share of all divorces. Sociologists Susan Brown and I-Fen Lin documented that the grey divorce rate doubled between 1990 and 2010, and it has continued rising. Grey divorces now represent approximately one in four divorces in the United States — a dramatic shift given that older cohorts were historically the most likely to remain in long-term marriages. Longer lifespans, women’s greater economic independence, and changed expectations about personal fulfilment in later life are cited as the primary drivers.

5. Divorce Statistics by Race and Ethnicity

Divorce rates vary substantially across racial and ethnic groups in the United States, a pattern that reflects both cultural factors and the confounding effects of socioeconomic disparities, since income, education, and age at marriage are all correlated with both race and divorce risk.

White Americans: Divorce rate sits close to the national average. The 10-year marriage dissolution rate for white women is approximately 32–35% according to NSFG data.

Black Americans: Have the highest divorce and separation rates of any major racial group in CDC and Census data. The 10-year dissolution rate for Black women is approximately 47–52%. Researchers consistently point out that this gap is substantially reduced — though not eliminated — when controlling for income, education, and age at marriage, suggesting that much of the disparity is driven by socioeconomic inequality rather than cultural difference per se.

Hispanic Americans: Divorce rates are below the national average and below the rates for white Americans. The 10-year dissolution rate for Hispanic women is approximately 27–30%. Researchers attribute this partly to stronger cultural and religious norms supporting marital stability, and partly to the fact that recent immigrant generations have lower divorce rates — rates that tend to converge toward the national average over subsequent generations.

Asian Americans: Have the lowest divorce rates of any racial group tracked by the Census Bureau, with 10-year dissolution rates approximately 15–20% for Asian women — roughly half the national average. This pattern is consistent across multiple data sources and over time.

Native American and Alaska Native populations: Have elevated divorce rates in Census data, broadly similar to or exceeding those of Black Americans, though with significant variation across tribal and geographic communities and with wider confidence intervals due to smaller sample sizes.

The racial gap in divorce rates substantially narrows when researchers control for income and education — meaning that the disparity reflects the unequal distribution of economic resources across racial groups rather than inherent cultural differences in attitudes toward marriage.

6. Education, Income, and Divorce Risk

Education is one of the most powerful predictors of divorce risk in the US data, and the relationship has strengthened over time. In the 1970s, divorce rates were relatively similar across educational groups. Today there is a substantial and widening educational gradient:

• Less than high school diploma: Approximately 46–50% 10-year dissolution rate
• High school diploma: Approximately 38–42%
• Some college: Approximately 35–38%
• Bachelor’s degree or higher: Approximately 20–25%

Income correlates similarly: households in the lowest income quintile divorce at roughly twice the rate of households in the top income quintile. Financial stress — particularly around housing, debt, and employment instability — is consistently identified as a major proximate cause of divorce. The ability to weather economic shocks together without catastrophic relationship strain is partly a function of having financial reserves, which higher-income couples are more likely to possess.

This means that the declining national divorce rate has been heavily driven by declining divorce among college-educated, higher-income Americans — while divorce rates among less-educated and lower-income Americans have remained stubbornly elevated. Sociologists call this the “diverging destinies” of American marriage: stable, low-divorce-rate marriage is increasingly concentrated among the already-advantaged, while relationship instability is disproportionately concentrated among those who can least afford its economic consequences.

7. Geographic Variation: Which States Divorce Most

Divorce rates vary substantially by state, with a consistent pattern that often surprises people who assume liberal coastal states would have higher rates.

Highest divorce rates are consistently found in: Nevada (which still counts many migratory divorces), Arkansas, Oklahoma, Wyoming, Alaska, and Kentucky. These are predominantly Southern and Mountain West states with lower median incomes, lower educational attainment, and higher rates of early marriage.

Lowest divorce rates are found in: Massachusetts, New York, New Jersey, Illinois, and Connecticut — states with higher educational attainment, later average marriage age, and higher median household income.

The irony — much discussed in demographic research — is that culturally conservative states, which tend to have the highest rates of marriage, also tend to have the highest rates of divorce. Researchers attribute this to the correlation between conservatism and early marriage, and early marriage and divorce risk.

8. What the Trends Reveal About Marriage in America

The overall picture of American divorce in 2025 is one of declining rates, demographic concentration, and widening class stratification. The national divorce rate is at its lowest point in roughly 50 years. But that average conceals a society increasingly divided between a college-educated, higher-income population for whom marriage has become more selective and stable, and a less-educated, lower-income population for whom relationship instability remains a significant source of economic and social vulnerability.

The practical implications of this data: divorce risk is not randomly distributed. It correlates strongly with age at marriage, educational attainment, income, and racial identity — the last of which functions largely as a proxy for the others given the racialised nature of economic inequality in the United States. Individual couples can meaningfully affect their own risk by choosing to marry later and ensuring economic stability before legal marriage — two factors consistently associated with dramatically lower divorce rates across every demographic group studied.

For more on the emotional landscape of separation, 12 signs your separated husband still loves you explores the complex dynamics that can persist after couples separate. And for context on the relationship patterns that precede partnership formation, why am I so single — 8 reasons offers an honest look at what actually prevents people from building lasting relationships in the first place.