Why does my Mac lose battery while it's asleep?

Updated 14 August 2026 · the measured numbers below come from six consecutive nights of one Apple silicon MacBook Pro

You closed the lid at 92% and opened it nine hours later at 74% — two points an hour, with nothing running. The forums will give you fifteen things to switch off; this page gives you the arithmetic to find out whether you have a problem at all — and, if you do, the name of what's causing it. macOS keeps that name in a log it never shows you.

First: is your overnight loss actually abnormal?

A sleeping Mac isn't switched off. It keeps memory alive and the network reachable — and that costs far less than people assume. Apple publishes the measurement in its own environmental reports: 0.22 W asleep for the 14-inch MacBook Pro (M4), 0.40 W for the 16-inch. On those packs it works out to roughly a third of a percentage point per hour, so pure sleep across a nine-hour night is around 3%. Add the scheduled wake-ups below and a night landing near 3–5% is the machine behaving.

One caveat before you use those rows: the same watt costs a different percentage on different Macs. A 13-inch Air carries about 54 Wh, a 16-inch Pro 100 Wh — so an identical draw shows up nearly twice as fast on the Air. If your Mac is at the small end, read the bands generously.

And if the Mac was warm in the morning or the fans had run, skip the division entirely: it wasn't asleep. Go straight to finding out for whom.

Do the division before you change any settings: points lost ÷ hours asleep. Most threads about overnight loss turn out to be normal sleep measured against an expectation of zero.

Your Mac wakes dozens of times a night — silently

This is the part nobody tells you. A sleeping Mac wakes on a schedule to do maintenance, then goes back to sleep, without the screen or a sound. On the machine we measured, over six nights:

Thirteen minutes a night of quiet housekeeping is the normal case — and it's the part that sits on top of the 0.22 W baseline, because a Mac that's awake for half a minute is using far more than a sleeping one. The abnormal case looks the same in the status menu and completely different in the log: one service waking the machine hundreds of times, or holding it awake for many minutes at a stretch.

Who's asking — the answer nobody prints

macOS records every scheduled wake and the task that requested it. On the Mac we measured, two thirds of the requests carried a name, and the biggest requester wasn't the search indexer everyone blames — it was the Calendar's travel engine, followed by generic maintenance, search heartbeats, DHCP lease renewals and disk-encryption health reports.

Then we repeated it on a second Mac and got a different answer entirely: three times as many wake-ups, a sixth as long each, and widgets rather than Calendar at the top. That's the real finding — the culprit is specific to your machine, so the only list worth acting on is your own.

Full numbers, method and the top requester table are in our night-wake study — including the command to run the same analysis on your own Mac in Terminal.

WattMate does this reading for you: for each night it says how long the Mac actually slept (not just how long the lid was closed — those are different numbers), how many times it woke, and which services asked. That works whether the night was on battery or on the charger.

The usual suspects, in the order worth checking

1. Something kept the Mac awake, not asleep. An external display, a hub, a connected iPhone or an app holding a "prevent sleep" assertion. If the machine was warm in the morning, start here.
2. Power Nap. On battery it lets the Mac check Mail and refresh Calendar and iCloud in its sleep; software updates and Time Machine backups wait for the charger. A real trade of battery for freshness — turn it off in System Settings → Battery → Options if you'd rather have the charge.
3. Wake for network access. Handy for remote access and Find My; on a busy network it adds wake-ups.
4. A misbehaving background service. The case where the log earns its keep: a specific requester appearing hundreds of times, or holding the machine awake for minutes. You can't guess this one — you have to read it.
5. Battery age. A pack with high cycles gives up percentage faster everywhere, including at night. Check health separately before blaming sleep.

Notice what's not on this list: reinstalling macOS, resetting the SMC on an Apple silicon Mac (Apple's own instruction there is just to restart — it resets itself), or the dozen ritual fixes that fill forum threads. Measure first, then change one thing.

What can't be measured honestly

The question everyone asks — "how much battery did those 124 wake-ups cost?" — has no honest answer, and we'd rather say so than invent one. macOS reports charge in whole percentage points; a 31-second wake doesn't move a whole point. Anyone giving you milliwatt-hours per wake-up is modelling, not measuring.

What is measurable: how long the Mac slept, how often it woke, for how long, who asked, and the total charge the night cost. That's what WattMate prints, and where a number would be a guess it says so instead.

Questions people ask

How much battery should a Mac lose overnight?

Roughly a third of a percentage point per hour while genuinely asleep, plus the scheduled wake-ups — call it 3–5% across a night on a mid-size pack, more on a small one. Much above that means it wasn't really asleep.

Why is my Mac warm after being closed all night?

Because it was awake, not asleep. Something either prevented sleep or woke it repeatedly — the power log names the requester, and that's the fastest route to the culprit.

Does closing the lid put a Mac fully to sleep?

Not entirely — scheduled maintenance wakes are by design, and every one on our test machine was silent. Normal in small doses; the log is what tells you whether your doses are small.

Is turning off Power Nap worth it?

If overnight charge matters more to you than waking up to fetched mail, yes. It's a genuine trade, not a bug fix — and you can measure the difference across two nights instead of trusting either side of the argument.

Related: the night-wake study, with the raw numbers · why charging says "on hold" · what to expect after the macOS 27 update