Short answer: snoozing is doing less damage than you were probably told, but it is also a symptom of something worth fixing. Two large studies published eighteen months apart reached opposite-sounding conclusions. Read both and the practical advice gets clearer than either headline suggests.
Almost everyone snoozes, and the pattern gives the game away
In May 2025, a team led by Rebecca Robbins at Brigham and Women's Hospital published the largest look at snoozing to date in Scientific Reports: more than 3 million nights logged by over 21,000 people using a sleep-tracking app.
- 55.6% of sleep sessions ended with the snooze alarm.
- On those mornings, people pressed it 2.4 times and spent 10.8 minutes snoozing.
- Around 45% of users snoozed on more than 80% of mornings, averaging roughly 20 minutes a day.
- Usage was highest in the United States, Sweden and Germany, lowest in Japan and Australia.
- It peaked Monday through Friday and collapsed on Saturday and Sunday.
That last line is worth sitting with. If snoozing were a purely biological need, it would not take the weekend off. What drops on Saturday is not fatigue. It is the alarm.
So is hitting snooze actually bad for you?
Here is where it gets awkward. In October 2023, Tina Sundelin and colleagues published two experiments in the Journal of Sleep Research and found essentially no short-term harm. Habitual snoozers lost about six minutes of sleep on snooze mornings. Thirty minutes of snoozing did not degrade cognitive performance, and on several tests it slightly improved it. No measurable effect on the cortisol awakening response, on mood, or on self-reported sleepiness.
That got reported almost everywhere as a green light. It is not quite one.
Robbins reads her own data the other way: "The snooze alarm disrupts some of the most important stages of sleep. The hours just before waking are rich in rapid eye movement sleep. Hitting the snooze alarm will interrupt these critical stages and typically only offer light sleep in between snooze alarms." A 2022 paper in the journal SLEEP by Mattingly and colleagues found the same signature in wearable data: snoozers showed lighter sleep in the hour before waking and higher resting heart rates than people who got up on the first alarm.
The two results are less contradictory than they look. Sundelin measured what happens on a single morning to people who already snooze. Robbins and Mattingly describe a habit repeated five days a week for years, in the REM-densest window of the night. Six minutes once is noise. Twenty minutes a day for a decade is a question nobody has run the trial to answer.
The finding nobody put in the headline
Buried in the 2025 data is the most useful result of the lot: people who went to bed later than usual snoozed more, and long sleepers snoozed more than short sleepers. Snoozing tracks bedtime, not sleep debt. Most people who do it have set an alarm they never intended to obey.
Robbins's recommendation is refreshingly blunt: "Set your alarm for the latest possible time, then commit to getting out of bed when your first alarm goes off."
Most of us do the exact opposite. We set the alarm half an hour early as insurance, spend that half hour in the worst sleep of the night, and open the day having already broken a promise to ourselves.
What you are actually fighting is sleep inertia
Sleep inertia is the fog between waking and full alertness. It slows reaction time, degrades short-term memory and impairs reasoning. For most adults it runs 15 to 60 minutes, and after real sleep loss it can stretch well beyond that. Work cited by NIOSH puts roughly 20 minutes as the time needed for cognitive performance to climb back to a level that reflects the sleep you actually got.
This is the part that reframes the whole problem. Inside that window you make the decisions that set the tone of the day: whether you train, whether you eat, whether you leave calmly or in a scramble. And you make them with a brain that is measurably impaired.
Lying still does nothing to shorten sleep inertia. Bright light, physical movement, cold water and tasks that demand genuine cognitive effort all do.
Why willpower is the wrong tool at 6am
The version of you that sets the alarm at 11pm is rested and reasonable. The version that hears it is neither. Asking morning-you to honor a decision made by night-you is asking the wrong party to perform.
Which is why every alarm trick that reliably works removes the decision instead of trying to strengthen it. Put the clock across the room. Sleep in your running clothes. Load the coffee machine before bed. None of them need motivation at 6am, because the rested version of you already spent it.
Phones quietly deleted the best of those tricks. A clock on the far wall forced you to stand up. A phone alarm dies under a thumb, often with no memory of it happening at all.
A few apps have rebuilt that friction on purpose. Risly is an anti-snooze alarm app for iPhone with no snooze button and no off switch: you pick a mission the night before, and the alarm keeps ringing until it is done. There are seven. Scanning a specific everyday object, your toothbrush or the coffee machine, forces you across the room. Chained mental arithmetic drags the prefrontal cortex online. Push-ups counted by the camera get your heart rate up. A printed QR code, or a barcode you already own on a shampoo bottle, becomes an off switch you have to walk to. A flash-grid memory task cannot be cleared on autopilot. Shaking hard and long enough is there for the mornings you want friction without leaving the bed yet.
Whatever tool you use, the principle holds: dismissing the alarm has to cost something a half-asleep brain cannot pay.
Five things that hold up
- Set one alarm, at the latest time that genuinely works. Then get up. This is the single change with the most evidence behind it.
- Keep the same wake time on weekends. Your circadian system responds to regularity more than to total hours in bed, and the 2025 weekday-weekend gap is exactly what regularity closes.
- Get bright light in the first half hour. It is the strongest signal available for shutting down melatonin production.
- Move before you negotiate. Sixty seconds of actual movement beats any amount of lying still deciding whether you feel ready. You will not feel ready. That is what sleep inertia is.
- Fix the bedtime, not just the morning. If you are wrecked at 6am, the honest cause is usually 12:40am. No alarm design compensates for chronic sleep debt, and none should be asked to.
The honest conclusion
If you snooze once, get up ten minutes later and feel fine, the evidence says relax. If you snooze for twenty minutes every weekday, wake up foggy and quietly dread your mornings, the button is not the disease. It is the most visible symptom of an alarm set at a time you never really agreed to.
Pick a wake time you can defend, set one alarm for it, and make getting out of bed the only thing that stops the noise. Then judge whatever system you land on by a single test: on the mornings you feel worst, does it still work?