SmartClock

How Much Sleep Did I Get? Count It Honestly

By Mark Fulton · 2026-09-25 · 11 min read

How Much Sleep Did I Get? Count It Honestly

To work out how much sleep you got, start with the time between lights out and your final wake-up, then subtract how long it took you to fall asleep and every stretch you lay awake in the night. Sleep = time in bed, minus time to fall asleep, minus time awake. In bed from 11 PM to 7 AM is eight hours in bed, but with 25 minutes to drift off and two wake-ups of 20 and 10 minutes, it's about 7 hours 5 minutes of sleep. That gap is why a night that "should" have been eight hours can feel like six. Divide the sleep by the time in bed and you get your sleep efficiency: here, about 89%.

Most sleep calculators answer a different question: what time to go to bed tonight. This page is about last night. It walks through the arithmetic, shows one night counted three ways, explains what sleep efficiency is and where the idea comes from, and is honest about what a watch or ring can and can't tell you.

Why is time in bed not the same as time asleep?

Because you're awake for some of it, and you rarely notice how much.

A night in bed has three parts. There's the time before you fall asleep, which sleep researchers call sleep onset latency. There's the sleep itself. And there's the time you spend awake after first falling asleep, whether that's a trip to the bathroom, a child calling, or the 3 AM stretch of staring at the ceiling. Clinicians call that last part wake after sleep onset.

Only the middle part is sleep. Subtracting bedtime from wake time counts all three, which is why the simple "11 to 7 is eight hours" math almost always flatters the night.

There's a fourth piece people forget: time in bed that was never an attempt to sleep. Scrolling your phone for 30 minutes before you turn the light off isn't sleep latency. It's just time in bed. A 2016 paper in the Journal of Clinical Sleep Medicine on measuring sleep efficiency makes exactly this point, arguing that reading, watching TV and other non-sleep activity before lights out shouldn't count against you, and that the clock should start when you actually try to sleep. We use that rule below. Start counting at lights out, not at "got into bed".

It also matters that some wake-ups are too short to remember. The NHLBI's page on sleep phases and stages notes that a sleep cycle restarts every 80 to 100 minutes, with four to six cycles a night, and that you may wake briefly between cycles. Brief wakings like that won't show up in your own count, and that's fine. The ones you remember are the ones worth subtracting.

How do you count the hours you actually slept?

Four numbers, one line of arithmetic.

  1. Lights out. The time you put the phone down and tried to sleep.
  2. Time to fall asleep. Your honest estimate. According to the Sleep Foundation's explainer on sleep latency, about 10 to 20 minutes is normal for most adults. If you don't know, use 15 and correct it over a few nights.
  3. Time awake in the night. Add up each wake-up you remember. A guess to the nearest five minutes is fine.
  4. Final wake-up. When you woke for good, not when you finally got out of bed. If the alarm went off at 7 and you'd been awake since 6:30, your final wake-up was 6:30.

Then:

Sleep = (final wake-up minus lights out) minus time to fall asleep minus time awake in the night.

That's the whole calculation. The reason it feels harder than it is: most people only remember numbers 1 and 4, and those are the two that make the night look longer than it was.

The worked example: one night, counted three ways

Here's a real-shaped night. Lights out at 11:00 PM, alarm at 7:00 AM. It took about 25 minutes to fall asleep. There was one wake-up at 2:10 AM that lasted around 20 minutes, and another near 5:30 AM of about 10 minutes.

How you count it Arithmetic Sleep you'd claim
Bedtime to alarm 11:00 PM to 7:00 AM 8 h 00 m
Minus time to fall asleep 8 h 00 m minus 25 m 7 h 35 m
Minus two wake-ups 7 h 35 m minus 20 m minus 10 m 7 h 05 m

Sleep efficiency for this night: 7 h 05 m ÷ 8 h 00 m = 425 ÷ 480 minutes = about 89%.

Two things stand out. First, 55 minutes disappeared without anything dramatic happening: a slow start and two ordinary wake-ups. Second, this still counts as a reasonable night. It clears the seven-hour mark, and the efficiency is solid. The same pattern starting at midnight would not.

Now suppose you also got into bed at 10:30 and read until 11. The honest count doesn't change, because the clock starts at lights out. But if you'd counted from 10:30, you'd have told yourself you were in bed eight and a half hours and felt even more cheated by the morning.

A quick log so the numbers are real

Memory is a poor sleep counter, especially for wake-ups. A few nights of notes beat a month of guessing. Keep it this short:

Field Example
Lights out 11:00 PM
Minutes to fall asleep (estimate) 25
Wake-ups (minutes each) 20, 10
Final wake-up 7:00 AM
Sleep 7 h 05 m
Efficiency 89%

Write it in the morning, not at 3 AM. If you do want to note the time of a wake-up in the night, a dim bedside display like the night clock shows it without the glare of unlocking a phone.

What is sleep efficiency?

Sleep efficiency is the share of your time in bed that you spent asleep. The usual formula is total sleep time divided by time in bed, times 100. The 2016 paper linked above calls it central to insomnia research and practice, and it's the number that drives a core part of cognitive behavioral therapy for insomnia (CBT-I).

In that therapy, a technique called sleep restriction uses efficiency as the steering signal. Per the same paper, participants are usually told to keep shortening their time in bed while efficiency stays below about 85%, to hold the schedule steady between 85% and 89%, and to add time back once efficiency reaches 90% if they still don't feel rested. Those thresholds are clinical working rules inside a supervised program. They're useful for reading your own number, but they aren't a diagnosis, and sleep restriction is not something to run on yourself on the strength of a blog post.

The denominator matters more than it looks. The paper works through a sample diary where the conventional formula gives 55% and a denominator that leaves out pre-sleep reading and post-wake lounging gives 64%. Same night, very different number. That's why our method starts at lights out and ends at your final wake-up: it measures how well you slept during the time you were trying to, not how long you lingered in bed.

Reading your own number, loosely:

Your efficiency What it usually means
90% or higher Most of your attempt was sleep. If you're still tired, the issue is likely total hours, not the quality of the time in bed.
85% to 89% The range sleep restriction programs aim to hold steady. A reasonable night.
Below 85% on most nights A lot of your time in bed is awake. Worth tracking for two weeks, and worth raising with a clinician if it persists.

How accurate are sleep trackers at this?

Better at spotting sleep than spotting wake, which is exactly the part this calculation depends on.

A 2021 study in the journal SLEEP tested seven consumer sleep trackers (four wearables and three bedside devices) against polysomnography, the full lab sleep study, in 34 healthy young adults over three nights. Every device was good at recognizing sleep when the person was asleep, with sensitivity of at least 0.93. They were much weaker at recognizing wake when the person was awake, with specificity ranging from 0.18 to 0.54. The devices also tended to do worse on nights with disrupted sleep. Most performed as well as or better than research-grade actigraphy, and sleep stage estimates were inconsistent.

In plain terms: lying still and awake can get counted as sleep. On a smooth night your tracker's total is probably in the right area. On a rough night, which is the night you most want an accurate number, it's more likely to be generous.

The American Academy of Sleep Medicine's position statement on consumer sleep technology takes the same careful line. It says that, given the lack of validation and FDA clearance, these devices can't be used to diagnose or treat sleep disorders, but that their data can be useful when brought into a proper clinical conversation.

So use the tracker for trends, not verdicts. If the watch says 7 h 40 m and your own count says 6 h 50 m, the honest count is often closer on a broken night. A "sleep score" is a company's blend of several measures, and it isn't the same thing as hours slept.

How many hours should the answer be?

For adults, the benchmark is seven or more. The joint consensus statement of the American Academy of Sleep Medicine and the Sleep Research Society says adults should sleep 7 or more hours per night on a regular basis, and that regularly sleeping less than 7 is associated with adverse health outcomes. It also notes that sleep need varies between individuals and that more than nine hours may be appropriate for young adults, people recovering from sleep debt and people who are ill.

Two practical readings of that:

  • The benchmark applies to sleep, not time in bed. Seven hours in bed is not seven hours of sleep. Using the worked example's losses, getting seven actual hours takes closer to eight hours between lights out and your final wake-up.
  • One night doesn't define your pattern. "On a regular basis" is the key phrase. Look at a week of honest counts, not a single bad Tuesday.

If you're still picking a bedtime to hit that number, what time should I go to bed has tables for every alarm from 5 to 9 AM, including how to correct for the time you take to fall asleep.

What should you change after a short night?

Mostly, not much, and not in a panic.

  • Keep the wake time. Sleeping in after a bad night shifts your body clock and can make the next night harder. A steady wake time is the anchor most sleep guidance starts from.
  • Don't go to bed much earlier to "make it up". More time in bed that you can't sleep through lowers your efficiency and gives the next night more room to go wrong. A slightly earlier lights-out is reasonable if you're genuinely sleepy.
  • Fix the biggest subtraction. If your count shows 45 minutes to fall asleep, that's the lever, not total bedtime. If it's a long 3 AM wake-up, that's the one to look at.
  • Separate tired from groggy. A foggy first half hour after the alarm isn't always a short night. It can be sleep inertia, which we covered in why you wake up groggy.
  • Know when it's not a timing problem. If most nights fall below seven hours despite enough time in bed, or you regularly take more than half an hour to fall asleep, or you're exhausted during the day, that's a conversation with a clinician. CBT-I is delivered by trained professionals for exactly this pattern.

Plan tonight so the count comes out right: enter your alarm on the sleep calculator and it gives bedtimes that count back in roughly 90-minute cycles with time to fall asleep built in. If your honest count says you take longer than that to drift off, move lights-out earlier by the difference. It's free, runs in your browser and needs no account.

FAQ

How do I calculate how many hours I slept?

Take the time from lights out to your final wake-up, then subtract how long it took to fall asleep and any time you were awake in the night. For example, lights out at 11 PM and awake for good at 7 AM is 8 hours. Minus 20 minutes to fall asleep and 15 minutes awake at 3 AM, that's 7 hours 25 minutes of sleep.

Does lying awake in bed count as sleep?

No. Resting with your eyes closed isn't sleep, and it doesn't count toward your total. It's worth knowing that brief wakings between sleep cycles are normal and often go unremembered, so you don't need to subtract those. Subtract the stretches you clearly remember being awake for.

Is my smartwatch sleep score accurate?

The total sleep figure is usually in the right area on a smooth night. A 2021 study of seven consumer trackers found they detected sleep well but were much weaker at detecting wake, and did worse on disrupted nights. That means a restless night can be overcounted. The sleep score itself is a manufacturer's composite, not a clinical measure. The AASM says these devices can't be used to diagnose sleep disorders.

How much sleep did I get if I slept from 11 PM to 6 AM?

At most 7 hours, and probably less. If it took you about 15 minutes to fall asleep and you were awake for around 20 minutes in the night, you got about 6 hours 25 minutes. That's under the seven hours or more recommended for adults. One night like that is fine. A week of them is a pattern worth changing.

Sources

SmartClock is a free set of alarms, timers, and world clocks. No sign-up, no ads, runs in your browser.