“What do you have to earn to be in the top 10%?” has no single answer, because the people it compares you with differ. Someone aged 25 to 34 earning $137,000 is level with the 90th percentile of their own age group and well below the 90th percentile of 45-to-54-year-olds. This guide reads the line off the Census Bureau's own tables for 2025 earnings, age group by age group, then by sex and by education, and checks the middle of it against a second government source. The same numbers are behind the salary percentile calculator, where you can put in your own pay.
The top 10% line, by age
The comparison group is full-time, year-round workers: those who usually worked 35 or more hours a week for 50 or more weeks in 2025 (the Census definition), men and women together. Earnings are wages and salaries plus profit from self-employment, before tax. Figures at $100,000 and over are rounded to the nearest $1,000 because the Census tables count everyone at $100,000 and over in one row and the spread above it is estimated.
| Age | Median | Top 25% starts | Top 10% starts | Top 5% starts |
|---|---|---|---|---|
| 25 to 34 | $61,020 | $90,400 | ~$137,000 | ~$178,000 |
| 35 to 44 | $74,910 | ~$118,000 | ~$185,000 | ~$247,000 |
| 45 to 54 | $75,500 | ~$122,000 | ~$194,000 | ~$265,000 |
| 55 to 64 | $71,500 | ~$114,000 | ~$183,000 | ~$248,000 |
Two things stand out. The median rises from $61,020 at 25 to 34 to $74,910 at 35 to 44 and then barely moves ($75,500 at 45 to 54, $71,500 at 55 to 64), so most of the climb happens in the first ten years. And the top 10% line is still rising a little when the median has stopped, to $194,000 at 45 to 54 (5% above the 35-to-44 line and 42% above the 25-to-34 one). The people near the top of an age group are the ones whose pay keeps growing for longer, through promotions, partnership or a business that has taken off. Past 55 the line slips back to $183,000, which does not mean pay falls for the same people. These are different people at each age, one year's snapshot, and not everyone at the top stays in a full-time job into their sixties.
If you include part-time and part-year workers (“everyone with earnings”) every line is lower, because those workers earn less over a year: the top 10% line for all ages is then about $154,000. The calculator shows both. Use the full-time line if you work a full-time job, since that is the like-for-like group.
Men and women
The same table split by sex shows a gap at every age, at the median and at the top-10% line alike. The figures compare all men and all women working full-time, year-round in each age group; they do not hold occupation, hours beyond 35 a week, experience or anything else constant, so they describe the gap as it is, not why it exists.
| Age | Men: median | Women: median | Men: top 10% starts | Women: top 10% starts |
|---|---|---|---|---|
| 25 to 34 | $63,250 | $56,940 | ~$148,000 | ~$124,000 |
| 35 to 44 | $81,560 | $65,840 | ~$202,000 | ~$160,000 |
| 45 to 54 | $82,440 | $66,250 | ~$219,000 | ~$161,000 |
| 55 to 64 | $80,960 | $61,710 | ~$209,000 | ~$149,000 |
At 35 to 44 a woman working full-time, year-round is paid a median of 81% of a man's, and the top-10% line for women is 79% of men's. The practical point for reading your own number: the same $160,000 that is the top 10% of women aged 35 to 44 is about the 83rd percentile of men that age, so it matters which group you are asking about.
By education
Education makes the biggest difference of the three cuts here. For people 25 and over working full-time, year-round:
| Highest education | Median | Top 10% starts |
|---|---|---|
| Less than a high school diploma | $41,400 | $82,400 |
| High school graduate (including GED) | $51,500 | ~$110,000 |
| Some college, no degree | $60,400 | ~$134,000 |
| Associate degree | $61,700 | ~$130,000 |
| Bachelor's degree only | $83,400 | ~$196,000 |
| Master's degree only | ~$101,000 | ~$225,000 |
| Professional or doctorate degree | ~$131,000 | ~$353,000 |
A high school graduate reaches the top 10% of their own group at about $110,000, a bachelor's degree holder at $196,000, a master's degree holder at $225,000 and a professional or doctorate degree holder at $353,000. Notice the reverse reading: $196,000 puts a bachelor's degree holder in the top 10% of their group, yet the same pay would rank about the 98th percentile among high school graduates working full-time (workers aged 25 and over in both groups): the group you are measured against matters as much as the figure. The professional and doctorate rows are the least certain in the table: they combine two columns of the Census table, and the groups are small (about 5.7 million people).
How sure these numbers are
Two checks, both visible in the open. First, the middle: the Bureau of Labor Statistics publishes its own medians for full-time wage and salary workers, from the same Current Population Survey but a different measure (usual weekly earnings, wage and salary workers only, second quarter 2026, aged 25 and over). Multiplying a week by 52:
| Aged 25 and over | Census 2025 median | BLS weekly x 52 | Difference |
|---|---|---|---|
| Less than a high school diploma | $41,400 | $41,756 | -0.8% |
| High school graduate (including GED) | $51,480 | $51,688 | -0.4% |
| Bachelor's degree only | $83,430 | $84,656 | -1.4% |
| Bachelor's degree or higher | $92,170 | $91,936 | +0.3% |
The medians agree within 2% in all four rows even though the two measures differ and sit a year apart. For all ages together the Census median is $66,650 against $65,052 from BLS. At the top the two part company: BLS's 90th percentile for all full-time wage and salary workers is $152,048 (weekly $2,924 x 52) against about $171,000 here, 13% higher. The Census figure counts profit from self-employment as earnings and the year's total, where BLS counts only wage and salary workers' usual weekly pay, so some difference at the top is expected; we cannot split it between the two. Treat the top-10% lines as “about”, and the true cut-off for wage and salary workers alone as somewhat lower.
Second, the method itself, described in full in the calculator's questions: below $100,000 it follows the Census Bureau's own interpolation and reproduces every printed median it was tested against; above $100,000 the spread is estimated from the Census Bureau's finer table, with a typical error of about a percentage point and up to about five in the least certain groups. And it is a sample survey: the median's own standard error is a few hundred dollars for the whole population and several times that for the narrow groups above.
Top 10% of people, top 10% of households
Do not mix the two. The top 10% of US households starts at about $259,000 of total money income, against $171,000 for a full-time worker's earnings, because a household adds up everyone in the home and counts income that is not pay. Two workers who each earn the all-ages median of $66,650 make $133,300 together, which is about the 68th percentile of all households but only about the 50th among married-couple households, most of which have two earners. The calculator ranks household income against households of the same type and age.
What “top 10%” leaves out
- Cost of living. The tables are national. $194,000 buys a different life in different places, and nothing here adjusts for it.
- Hours and benefits. A percentile of pay says nothing about hours beyond 35 a week or employer benefits such as health insurance and retirement matches, which are not earnings in these tables.
- One year. The percentile is for one calendar year of earnings. Bonuses, a business's good year or a stretch of overtime can move someone across the line for a year and back.
- Who is counted. People in institutions and people with no earnings are not in the full-time group; the “all adults” option in the calculator includes the second group.
Not financial advice. These are descriptive statistics for 2025; comparing pay with a group says nothing about what any one person should earn or accept.
Sources
- U.S. Census Bureau — Table PINC-03, people by total money earnings in 2025, work experience, age, sex and educational attainment (CPS ASEC 2026): the counts every figure here is built from
- U.S. Census Bureau — Table PINC-11, income distribution to $250,000 or more (CPS ASEC 2026): the shape used above $100,000
- U.S. Census Bureau — Table HINC-02, households by total money income, age of householder and type of household (CPS ASEC 2026)
- U.S. Census Bureau — CPS ASEC 2026 Source and Accuracy Statement and glossary: universe, 'year-round full-time worker', linear interpolation for medians
- Bureau of Labor Statistics — Usual Weekly Earnings of Wage and Salary Workers, Table 5 (second quarter 2026, released July 2026): quartiles and deciles of weekly earnings by education, used as an independent check