O! Mental Health Blog

Why Do You Get a Second Wind at 11pm?

Why Do You Get a Second Wind at 11pm?

Why Do You Get a Second Wind at 11pm?

Your brain is done by late afternoon. The four o'clock call takes everything you have, you tell yourself you will be in bed by eleven, and then at around half past ten something switches back on. You are alert. The thinking you could not do at three suddenly feels available. You reopen the model, you answer the email you had been avoiding, and at one in the morning you are exhausted but cannot switch off.
This is not a niche complaint. In a survey of 2,007 US adults commissioned by the American Academy of Sleep Medicine and fielded in June 2025, 44 percent said they always or often have difficulty falling or staying asleep, and 26 percent said they prioritise phone time over the recommended amount of sleep. Most operators file the late surge as a scheduling discovery: late night is when I do my best work. That conclusion rests on a measurement error, and an expensive one, because the window arrives at the moment your system was about to let you recover.
The second wind is not extra capacity. It is the circadian alerting signal reaching peak output in the two to three hours before your own melatonin rises, offsetting sleep pressure that has been accumulating since you woke. That pressure is not cleared during the window. It is masked, and it is still there at one in the morning.
Why does your alertness rise just before you are supposed to sleep?
Two systems set how awake you feel. One is homeostatic: adenosine and related signals build the longer you have been awake, which is why the middle of the afternoon is hard. The other is circadian: the suprachiasmatic nucleus runs a wake-promoting signal that does not fade through the evening but climbs.
Sleep science has named the result twice. Lavie called it the forbidden zone for sleep in 1986, a stretch roughly two to three hours before habitual bedtime when falling asleep becomes unusually difficult. The circadian literature calls the same interval the wake maintenance zone. It is the one point in the 24 hours when the alerting signal runs at full output against near-maximum sleep pressure. Then melatonin onset arrives, the signal withdraws, and everything that was held back lands at once.
What does the second wind actually improve?
This has been measured. In a Berlin constant-routine study (de Zeeuw and colleagues, Scientific Reports, 2018), twelve healthy men were kept awake for 40 hours in light below 5 lux with hourly cognitive testing, and the wake maintenance zone was defined as the three hours before dim-light melatonin onset, which averaged 21:17.
Performance did improve inside the window, but narrowly. Go/No-Go accuracy improved on the first evening, lapses on a sustained-attention test fell on the second, and the drowsiness signature in the EEG dropped. Higher executive function did not move at all: working memory and abstract reasoning showed no wake-maintenance-zone benefit. Subjective sleepiness did not drop either, while the objective markers did.
So the surge buys you vigilance. It does not buy you judgement, and it does not announce itself honestly. Feeling sharp at eleven at night is not evidence that capacity has returned, because the felt signal and the measured signal are not tracking each other here. You are reading an arousal state as a performance state, the same interpretation gap that makes a green recovery score feel like permission. It is why the late evening is a nervous-system-first question rather than a discipline one.
Is the surge a gate that locks you out of sleep?
Here the popular version overshoots. If you have read that missing your window condemns you to hours of lying awake, the adult data does not support it.
Monterastelli and colleagues (Frontiers in Sleep, 2024) gave 44 adults aged 31 to 46 and 44 adolescents repeated nap opportunities across the evening, with the forbidden zone defined as the two hours either side of dim-light melatonin onset. Mean onset in the adults was 21:29. Inside that window the adults still fell asleep in 9.2 minutes before onset and 7.0 minutes after, and still slept 69 and 90 minutes of a 120-minute opportunity. The adolescents took roughly twice as long. Sleep was measured by wrist actigraphy rather than in-lab recording, so treat the latencies as approximate.
By your thirties and forties, then, the gate is modest. What keeps you up past it is mostly what you do inside it: light, screens, and one more decision.
What can you measure without a lab?
You cannot sample melatonin at home, and you do not need to. Two observations are enough.
First, the clock time the surge arrives. Note it for seven evenings to the nearest fifteen minutes. For most adults it clusters, and that cluster is a workable proxy for their own pre-melatonin window.
Second, your time to fall asleep on the nights you work through the surge against the nights you stop. Device estimates are crude, so use the honest version: the time you put the light out, and the last time you remember. If the worked-through nights cost forty minutes and the stopped nights ten, that is the whole argument, measured on you.
What is the minimum useful protocol?
Seven minutes, once: take your week of notes, find your own surge time, and set one hard stop fifteen minutes before it. Not a bedtime. A stop on new input.
Two minutes, nightly: at that stop, write the thing you were about to start on one line and close the laptop. The surge is very good at making an unresolved decision feel both urgent and solvable, and writing it down removes the usual reason the window extends.
Ten minutes, nightly: drop the light level in the last stretch of the evening. Current circadian guidance puts the practical target under 10 lux in the three hours before sleep, which in a normal room means lamps rather than overheads. Dimming does not remove the alerting signal, but it stops you adding to it.
Zero minutes, and the one that pays most: stop using alcohol to close the window. It shortens the time to fall asleep and takes the back half of the night in exchange.
When should you stop interpreting and speak to a doctor?
A surge that reliably arrives at two or three in the morning rather than late evening is not a second wind, and the timing is worth raising with a clinician. Snoring with witnessed breathing pauses, or sleep that never leaves you refreshed however long you spend in bed, warrants an assessment for sleep apnoea. Unexplained fatigue lasting weeks, or persistent low mood or loss of interest in things you normally care about, belongs with a qualified professional rather than a protocol. And if closing down the evening takes more alcohol or a sedative than it used to, that escalation is itself the signal.
Where does this leave you?
With the surge reclassified from opportunity to information. It is evidence that your clock is working. What it costs is the recovery you were ninety minutes away from, and repeated across a quarter that stops being a late night and becomes hidden biological debt. Protecting the window is one of the cheapest items in your performance infrastructure: it will not raise the biological ceiling, but it stops you trading recovery capacity for vigilance you cannot use on hard problems anyway. The physiological erosion is slow, which is why it shows up first as a change in patience rather than in sleep.
If you want a wider read than one evening habit can give you, the free self-assessment at osapiens.expert/how-i-feel maps where your system is currently carrying load across sleep, stress, movement, nutrition and recovery. It takes a few minutes and leaves you a baseline to measure drift from.
Written by Vladislav Andreev, mental health educator and executive coach, founder of O!Sapiens. Education, not medical advice.
This article is educational and is not a substitute for medical advice, diagnosis or treatment. If you have persistent symptoms or concerns about your health, speak with a qualified physician or licensed professional.