Daily Life · Humanity estimated
How many people are sleeping right now?
Explore how many people are asleep around the world, with a changing estimate, a daily curve and a separate estimate for India.
people, approximately
36.5% of the world population in this model
Model sensitivity (not a confidence interval): 2.52B–3.56B
≈ Modelled estimate · updated continuously
Typical estimate · modelled
Typical 24 Hours
The interactive 24-hour curve loads with the live model.
This curve shows a typical modelled day, not historical observations.
See how many people are awake right now using the same world population model.
How many people are sleeping in India right now?
The model estimates about 501M people sleeping in India. The model uses India’s local time and a sleep-duration input from the 2024 India Time Use Survey. The daily survey average informs the model, while the hourly pattern and under-15 contribution involve separate assumptions.
How this sleeping estimate works
Σ (country adult population × survey-based local probability + child population × assumed local probability)
US ATUS 2025: 9.03 hours for age 15+. India 2024: 546 minutes for age 6+, used as an explicit age-group proxy for India age 15+. Under-15 sleep retains the earlier 480-minute assumption. US hourly sleep includes naps and sleeplessness; it is a proxy for sleep, not physiological observation. Country population is interpolated between UN projection dates. IANA local time and regional weekend conventions drive the simulation. Borrowed timing and behavioural coefficients are not local measurements.
Is a third of the world asleep right now?
Not necessarily. Eight hours of sleep in a 24-hour day would mean one third asleep on average in a simplified model. At a particular moment, the share depends on where people live and their local sleeping hours. The counter and daily curve show that changing overlap.
Use the daily high and low above to compare the same model through the day. The India section uses the same instant, so you can compare its local sleep window with the worldwide estimate.
Direct answers
Questions people ask
How is sleep estimated worldwide?
Country populations and age structures weight national daily-duration anchors and borrowed local-hour profiles. Children’s sleep, survey scope differences, and countries without a matched sleep survey remain modelling assumptions.
What changes this estimate?
The estimate rises when populous regions move into local night-time and falls as those regions reach morning. Work schedules, age, culture and irregular sleep are important sources of uncertainty.
A proxy for sleeping, including naps. The US source category also includes sleeplessness, which the model cannot separate. Adult survey values and under-15 assumptions are documented separately.
For the other side of the same model, see how many people are awake right now. The two estimates use the same current world population; they are not independent surveys.
Baseline sources
US ATUS 2025: 9.03 hours for age 15+. India 2024: 546 minutes for age 6+, used as an explicit age-group proxy for India age 15+. Under-15 sleep retains the earlier 480-minute assumption. US hourly sleep includes naps and sleeplessness; it is a proxy for sleep, not physiological observation.
- United Nations — World Population Prospects 2024
- United Nations Statistics Division — Guide to producing time-use statistics
- U.S. Bureau of Labor Statistics — ATUS Table A-3 (2025)
- India MoSPI — Time Use Survey 2024
- National Bureau of Statistics of China — Third National Time Use Survey
- Eurostat — Harmonised European Time Use Survey 2020 metadata
- UN WPP 2024, original demographic workbook (GEN/F01/Rev.1)
- IANA Time Zone Database — civil time zones
- BLS ATUS 2025 — primary-activity daily hours (age 15+)
- BLS ATUS 2025 — weekdays, weekends and holidays (Table 2)
- World Bank / UN WPP — population ages 0–14, 2024
Status: National survey with explicit geographic/age proxies · Data vintage: Behaviour 2024–2025; population WPP 2024 projections; reviewed 8 September 2026. Population and clock review: 8 September 2026. Activity-specific limitations are stated above.
Details
When would the largest share of the world be asleep?
There is no single worldwide bedtime. The model combines local sleeping patterns with population weights, so the global peak depends on which large populations are in their sleep window at the same UTC moment. Read the peak from the modelled 24-hour curve above, not as a directly observed record.
The underlying diary category and assumptions matter as much as the clock. This model includes naps and cannot separate all sleepless time in the survey category. A difference between two hours describes the model’s distribution of sleep, not a change in sleep quality or a measurement of individual sleep needs.
How does reported sleep time differ on weekdays and weekends?
In the 2025 US time-use survey, the sleeping category averaged 8.76 hours on weekdays and 9.68 on weekends and holidays for people aged 15 and over. That difference is about 55 minutes in the diary category. It includes naps and sleeplessness, so it is not a measurement of 55 extra minutes physiologically asleep, a global pattern, or a target for an individual. BLS: 2025 American Time Use Survey, table 2.
Two sleep schedules that do not overlap
Illustrative example — not measured world data.
In a simplified group, one person sleeps during one eight-hour interval and another during a different, nonoverlapping eight-hour interval. Together they contribute 16 sleep-hours, but never two simultaneous sleepers. This invented schedule explains why total sleep time alone does not determine the global count at a chosen instant.
Is one third of the world always asleep?
No. Eight hours divided by a 24-hour day gives one third as a simplified daily average. It does not say that one third is asleep at every moment. The existing model distributes sleep across local hours and weights regions by population, so its simultaneous total rises and falls.
Are “asleep right now” and “sleeping right now” different counts?
They refer to the same estimate on this site. Use the number above for the current modelled moment and the daily curve for comparisons across hours. Opening a second page or refreshing does not create a separate population.
How many hours do people actually spend sleeping?
The US 2025 time-use table reports 9.03 hours per day in its sleeping category for people aged 15 and over. That category includes naps and sleeplessness, so it is not a medical measurement of time asleep. It is also a US survey result, not a global sleep duration or a recommendation. BLS: 2025 time-use table and sleeping footnote.
Why does the country distribution matter for sleeping right now?
A clock tells you the local hour, but population tells you how much that hour contributes to the world total. A sleep window covering a densely populated region contributes more people than the same window over a sparsely populated region. The model weights those populations rather than counting the dark half of a world map.
Can the daily sleep curve predict tonight’s exact number?
No. The curve applies typical schedules and documented assumptions, not observations of tonight’s behaviour. A festival, night shift, travel disruption or unusual heat may change real sleep without changing the baseline. Use the curve to understand the model’s timing, not as a forecast of measured sleep for a particular night.
See the complementary awake estimate
A snapshot of sleep versus a daily average
For a worked example, imagine 100 people who each sleep eight hours, but at different times. They contribute 800 sleeping hours across one day: an average of 33.3 people asleep, not necessarily 33 people at noon or midnight. The world counter adds geographical population weights and local schedules. Read its current number for this instant and its curve for the typical variation; neither is a direct observation of bedrooms.
Comparing sleep estimates from different websites
First check whether both figures describe the same timestamp and population. Then compare their treatment of children, naps, shift workers and local time. A fixed fraction of world population and an hourly regional model can differ substantially even when both use plausible daily durations. Extra digits show the calculation output, not extra certainty about individual sleepers.