The average American age 15 and over logs about 9 hours of sleep a day, spends just over 5 hours on leisure and works 3 hours. That last number looks wrong until you see how it is built, and nearly every headline figure from the American Time Use Survey works the same way. The survey is the most detailed record the United States keeps of where its hours go, and the Bureau of Labor Statistics published its latest edition, covering 2025, on June 25, 2026.
Its premise is simple: whatever time is, every respondent has exactly 24 hours of it per day, and the survey sorts those hours into categories that always add up to 24.00. The numbers below come from the 2025 release. Reading them correctly depends on knowing how a single day in a single person’s life becomes a national average.
How a phone interview becomes a national day
The survey is run by the Census Bureau for the BLS and has collected data continuously since January 2003. Its sample comes from households that have just finished their eighth and final interview for the Current Population Survey, the monthly survey behind the unemployment rate. From each household, one person age 15 or over is chosen at random and interviewed once by phone. In 2025, about 6,100 people took part.
Each respondent walks an interviewer through the previous day, activity by activity, from 4 a.m. to 4 a.m. The diary follows the respondent’s own clock, whichever US time zone they live in. Respondents are assigned the day they report on in advance: 10% report on each weekday and 25% on each weekend day. The oversampling gives the survey enough weekend diaries to measure them properly, and weighting then restores the real mix of five weekdays to two weekend days.
Two design rules shape every result. The survey records only the main activity at any moment, so watching TV while folding laundry counts as one or the other, not both. The only exception is secondary childcare, meaning time with a child under 13 in your care while doing something else. And every major category includes related travel: driving to work counts as work, driving to the gym counts as leisure.
The 2025 edition also has a gap. The federal government shutdown suspended all survey operations from October 1 through November 12, 2025, so no diaries exist for September 30 through November 11. The BLS reweighted the rest of the fourth quarter to stand in for the missing days and states that the effect on the estimates cannot be quantified, though it notes that time use has historically shown little seasonality.
The average American day in 2025
These are averages across everyone 15 and over, on all days of the week, including people who did not do the activity at all. The last column shows the share who did it on an average day.
| Activity | All (hours/day) | Men | Women | Share who do it on an average day |
|---|---|---|---|---|
| Personal care, including sleep | 9.80 | 9.59 | 10.00 | 99.9% |
| of which sleeping | 9.03 | 8.93 | 9.13 | 99.9% |
| Leisure and sports | 5.16 | 5.58 | 4.76 | 95.1% |
| of which watching TV | 2.61 | 2.81 | 2.42 | 74.3% |
| Working and work-related activities | 3.32 | 3.82 | 2.84 | 41.5% |
| Household activities | 1.99 | 1.58 | 2.38 | 80.9% |
| Eating and drinking | 1.21 | 1.24 | 1.18 | 96.9% |
| Purchasing goods and services | 0.69 | 0.61 | 0.77 | 39.7% |
| Caring for household members | 0.49 | 0.36 | 0.61 | 22.0% |
| Educational activities | 0.44 | 0.41 | 0.47 | 8.5% |
| Organizational, civic and religious | 0.28 | 0.25 | 0.31 | 13.6% |
| Other, not elsewhere classified | 0.27 | 0.28 | 0.27 | 20.4% |
| Phone calls, mail and email | 0.19 | 0.13 | 0.25 | 19.2% |
| Caring for people outside the household | 0.15 | 0.14 | 0.16 | 7.6% |
Source: BLS, American Time Use Survey, 2025 annual averages (Table 1). Hours are decimal, so 2.61 hours is 2 hours 37 minutes.
Why “3 hours of work” is not a mistake
The population average of 3.0 hours of work per day includes teenagers, retirees, people between jobs, and every Saturday and Sunday. On an average day, only 39.6% of people age 15 and over did any work at all. Those who did worked 7.65 hours.
Narrow the group and the familiar workday appears. Full-time employed people worked 8.1 hours on the days they worked: 8.5 hours on an average weekday and 5.5 on a weekend day. The weekday-weekend split is sharp: 81% of employed people worked on an average weekday, against 30% on an average weekend day.
Weekend work also follows education. Among people who worked on an average weekend day, high school graduates with no college put in 6.4 hours, while workers with a bachelor’s degree or higher put in 4.0. The survey doesn’t explain the gap, but the pattern fits a split between full weekend shifts in retail, health care and hospitality and a few hours of catch-up work at home. That is our interpretation, not a finding the BLS reports.
The full-time weekday figure of 8.5 hours is also above the 8-hour day built into the standard 2,080-hour year. The two numbers measure different things: the survey counts hours actually spent working on days people worked, while 2,080 is a payroll convention.
Nine hours of sleep, and what the number really measures
The sleep figure is the one most often misread. At 9.03 hours, it looks far more generous than the common warning that Americans don’t sleep enough. The difference is in the definition: the survey’s sleeping category explicitly includes naps and spells of sleeplessness. It captures the time people report spending asleep or trying to sleep, not physiological sleep measured by a device.
The average also includes teenagers. People ages 15 to 19 logged 10.9 hours of personal care a day, the highest of any age group, which pulls the national figure up. The weekend effect is visible too: 8.76 hours of sleep on weekdays and 9.68 on weekends and holidays, a gap of 55 minutes.
Television still takes half of all leisure
Leisure and sports averaged 5.16 hours a day, and television took 2.61 of them, almost exactly half. Three in four people (74.3%) watched on an average day, and those who did watched for 3.51 hours. On weekends and holidays, the population average rose to 3.27 hours.
Americans age 15 and over spend 2 hours 37 minutes a day watching television, more than housework, cooking and primary childcare combined (1 hour 38 minutes). worldtimedata
Over a year, 2.61 hours a day adds up to about 953 hours, by our arithmetic, close to half of a 2,080-hour full-time work year. The main-activity rule cuts both ways here: TV playing in the background while someone cooks counts as cooking, while dinner eaten in front of a show that the respondent names as the main activity counts as TV.
The ten-year trend shows where the rest of leisure has moved. In 2015, 38% of people spent some time socializing and communicating on an average day, for an average of 41 minutes. By 2025, that fell to 30% and 35 minutes. Over the same decade, time spent playing games or using a computer for leisure, a category that includes social media, rose from 25 to 37 minutes a day. Teenagers ages 15 to 19 led every age group at 1.5 hours.
Exercise is the smallest slice. Sports, exercise and recreation averaged 0.34 hours, about 20 minutes a day. That average hides a split: only 23.4% of people did any on an average day, and those who did spent close to an hour and a half.
Working from home settled at about one in three
The survey has tracked where people work since it began, which makes it one of the few consistent measures of remote work from before the pandemic to now. The measure counts anyone who did some or all of their work at home on days they worked, so a hybrid employee and an office worker answering email at night both count.
| Year | Employed people doing some or all work at home on days worked |
|---|---|
| 2003 | 19% |
| 2019 | 24% |
| 2021 | 38% |
| 2022 | 34% |
| 2023 | 35% |
| 2024 | 33% |
| 2025 | 35% |
Source: BLS American Time Use Survey releases for each year. The BLS did not produce annual estimates for 2020 because the pandemic disrupted data collection.
The share crept up from 19% in 2003 to about 23% by 2013, then barely moved until 2019. It jumped during the pandemic, eased back and has held at about a third since 2022, still well above the pre-pandemic level. In 2025, 70% of employed people did some or all of their work at their workplace, so a meaningful share did both on the same day. Employed women were more likely to work at home than men (38% versus 31%). Education is the biggest divider: among workers 25 and over, 51% of those with a bachelor’s degree or higher did some work at home, compared with 19% of those with a high school diploma and no college.
Housework, childcare and the gap between men and women
On an average day, 81% of people did some household activity, such as housework, cooking, lawn care or paperwork. The gap by sex shows up in both participation and duration: 87% of women and 75% of men did household activities on an average day, and on those days women spent 2.8 hours against 2.1 for men. Housework alone, meaning cleaning and laundry, shows the widest gap: 47.9% of women did some on an average day, compared with 23.8% of men.
Household time grows with age, from 43 minutes a day for people ages 15 to 19 to 2.7 hours for those 65 and over. Retirement shifts hours into the home rather than out of it.
Childcare is where the survey’s two-level design matters most. Adults living with a child under 6 spent 2.3 hours a day on primary childcare, such as feeding, bathing, reading or driving children somewhere. Within that group, women averaged 2.8 hours and men 1.7. But adults with a child under 13 also spent 5.1 hours a day on secondary childcare, meaning children were in their care while they did something else, most often leisure (1.9 hours) or household tasks (1.2 hours). Weekends flip the balance: parents of young children did more primary childcare on weekdays (2.4 hours versus 2.0) but far more secondary childcare on weekend days (7.7 hours versus 4.2).
Ages 35 to 44 have the least free time
Leisure follows a U-shape across life. People ages 35 to 44 averaged 3.89 hours of leisure a day, the lowest of any age group. They also logged the most work (4.83 hours on average across all days) and the most care for household members (1.23 hours). At the other end, adults 75 and over averaged 7.39 hours of leisure a day, nearly twice as much. Whether those extra hours feel faster or slower is a question a time diary cannot answer; it records clock time only.
Young children are closely linked to that squeeze. Employed adults with no children under 18 at home had 4.5 hours of leisure a day; employed adults living with a child under 6 had 3.2, about 1 hour 20 minutes less. The survey shows the association, not the cause, since age, job type and children overlap.
What the survey cannot tell you
The survey is built for averages across a large population, and its limits follow from that design. Each person reports a single day, so it cannot show how any individual’s week varies. It relies on recall of yesterday rather than tracking devices, which makes it better at capturing how people describe their time than at measuring it to the minute. The main-activity rule, described above, undercounts anything done in the background. And the sample has shrunk from about 9,400 respondents in 2019 to about 6,100 in 2025, which widens the margin of error on small subgroups; the BLS suppresses estimates that do not meet its reliability standards and generally tests differences at the 90% confidence level. For comparisons across years, the safest approach is to use the published tables from each year’s release and to treat changes of a few minutes as noise unless the BLS itself highlights them.
