In 2025, about 7% of opposite-sex couples living together had a male partner at least 10 years older. Census counted 5.1 million such couples and 1.3 million couples with a female partner at least 10 years older.
On this page
How it compares with smaller gaps
In the same 2025 Census study, more than one-third of opposite-sex couples living together were within one year of each other. About 7% had a male partner at least ten years older. Those are different parts of the age-gap distribution, and the published summary does not give every interval in between. Do not turn either figure into a rule for how an individual relationship will work.
The ten-year cutoff includes couples with a gap of ten years and couples with a wider gap. It does not tell us how many pairs are exactly ten years apart. Someone looking for a figure about an eight-year or twelve-year difference cannot read that value from this headline. The close-age measure has a different boundary at the other end of the distribution. Many couples fall between those two published descriptions, and the short Census summary does not break them into single-year steps.
The comparison works because both statements come from the same 2025 Current Population Survey description of opposite-sex couples sharing a home. If you bring in Pew's 2022 marriage-only age-gap results, you change both the year and the marital-status group. Those results can add context, but they should sit in their own row instead of being spliced into a single distribution.
Which ten-year gap?
The table may be wider than the screen. Focus it and use the arrow keys to reach every column.
| Age pattern | Published result | Reading limit |
|---|---|---|
| Within one year | More than one-third | Close-age reference group |
| Man at least 10 years older | About 7%; 5.1 million couples | The 7% describes this direction |
| Woman at least 10 years older | 1.3 million couples | No share given in the cited summary |
Percent scope: The two ten-year rows count opposite directions. The published 7% share belongs only to couples with the man older.
How to read it: Couples living apart are outside the household survey. These counts do not measure relationship duration or satisfaction.
Download the table (CSV) ↓Who is included
The 2025 Current Population Survey covers opposite-sex married and unmarried partners in the same household. Census counted about 5.1 million couples with a man at least ten years older and 1.3 million with a woman at least ten years older. The 7% share refers to the first direction only. It does not include the 1.3 million in its numerator, so it should never be presented as the share for all ten-year gaps.
The source counts a couple when both partners are identified in one home. People dating while living separately are outside this pair-based estimate. Same-sex couples are also outside the opposite-sex comparison used for these directional results. That restriction is about the published measure, not about whether such couples can have large or small age gaps. A survey of all relationships would need to identify a wider set of pairs and record both ages.
The count and percentage have different jobs. About 5.1 million tells us the estimated number of observed couples with the man ten or more years older. About 7% tells us what fraction of the observed opposite-sex co-resident couple group those couples represented. A large count can coexist with a modest percentage in a large population. Do not treat the 5.1 million as a count of people; each unit is a couple.
What happens when the woman is older
The Census Bureau separately counted about 1.3 million opposite-sex co-resident couples with the woman at least ten years older in 2025. This is a real part of the age-gap distribution, though the article's published 7% share belongs only to the man-older group. By our addition of the two rounded counts, about 6.4 million observed couples had a partner at least ten years older in one direction or the other. That total is an arithmetic summary of the published rounded counts, not a new Census table value.
The man-older count is about 3.9 times the woman-older count by our division of 5.1 million by 1.3 million. The underlying counts are rounded, so the ratio is approximate. It tells us which directional pattern was more common in these observed households. It does not measure which age gaps people prefer, who proposed the relationship, or whether either type of couple is more satisfied. Those questions were not answered by the two count figures.
It is tempting to turn 1.3 million into a percentage by dividing it by an assumed national couple total. The short Census summary does not publish the denominator beside that row. A responsible share needs the correct weighted group total from the same survey, not a household count from another year or a different study. Until that is in hand, quote the count for the woman-older direction and keep the 7% beside the man-older direction only.
A ten-year difference is not a single life stage
Imagine one hypothetical couple aged 22 and 32, and another aged 62 and 72. Each pair has a ten-year difference. The Census category treats them alike for the age-gap threshold, while their work, family, and health circumstances could be very different. These examples are arithmetic, not descriptions of people in the survey. They show why a threshold cannot answer a personal question about whether two lives fit well together.
Age as recorded in whole years also simplifies exact birth dates. A pair's stated ages at interview can differ by ten even when the exact number of months between birthdays falls on either side of a round anniversary. Our age-gap tool can calculate the date difference for two birth dates. That personal result is separate from the survey's grouped age estimate. Use the tool to answer 'what is our gap?' and the Census result to answer 'how often did this broad pattern appear among observed households?'
The figure offers no direct evidence about how relationships with a ten-year gap begin, how long they last, or how partners share power. The Current Population Survey observes households at a point in time. A relationship that ended earlier is not in the co-resident couple group, and a current pair may change later. Neither a concern nor a promise about an individual relationship can be derived from this prevalence result.
Use the statistic with its direction attached
For a clear citation, say: 'About 7% of opposite-sex couples living together in 2025 had a male partner at least ten years older, according to the Census Bureau.' If the question is about women ten years older, use the separate count of about 1.3 million and avoid assigning it the 7%. If the question is about every kind of couple or all people dating, neither estimate fully covers the population. The population and direction are part of the answer.
If you want context for a smaller gap, compare the Census within-one-year result on this page or read our close-age guide. Keep the definitions in view: within one year and at least ten years are different thresholds, and the published summary leaves the middle unspecified. A useful comparison can be modest. It can show that ten-year man-older gaps were less common than close-age pairs in this observed group without pretending to know every interval between them.
