Women's sleep

Postpartum insomnia: awake when the baby is finally asleep

The baby went down two hours ago and you are still staring at the ceiling. That is not newborn sleep deprivation. It is a different problem, and it has a different fix.

September 2, 20265 min read

Last reviewed September 2, 2026

A night broken into four separate stretches, held at the morning end by one bright fixed line.

Broken sleep after a baby is expected. Lying awake while the baby is asleep is not, and the difference matters more than almost anything else you will read about postpartum sleep.

One of those is a schedule problem imposed from outside. The other is insomnia, and it does not resolve on its own when the baby starts sleeping through.

The test that separates them

Sleep deprivation and insomnia feel similar from the inside and behave in opposite ways.

Sleep deprivation: you fall asleep the moment you get a chance. In the car. Mid-sentence. The constraint is opportunity, not ability.

Insomnia: the baby went down two hours ago, your partner has the monitor, the house is quiet, you are exhausted - and you are awake. Given the opportunity, you cannot use it.

The second one is the one this article is about. It is the same mechanism as being tired all day and wide awake at bedtime: sleep pressure is high, but arousal is higher, and arousal wins.

It is also more common than the silence around it suggests. In a 2026 study that used structured diagnostic interviews rather than questionnaires, 13% of women met full criteria for insomnia disorder at six weeks postpartum, and about a quarter had insomnia symptoms. Of the women with symptoms, almost none had been evaluated for it by a clinician.

Why it starts and then keeps going

For months, waking at 2am was correct. Your body learned it. Lighter sleep, faster arousal, listening for a sound - all of that was adaptive and none of it switches off on schedule.

Then the compensating starts, and the compensating is what turns a phase into a condition. You go to bed at 8:30pm because you might get a long stretch. You stay in bed at 5am because you might drift off again. You nap when you can.

Every one of those is sensible in the moment. Together they do a specific thing: they stretch the hours you spend in bed far past the hours you actually sleep, which is the single most reliable way to teach a brain that bed is a place where you lie awake. That is the same trap described in sleep restriction therapy, arriving by a different route.

What "sleep when the baby sleeps" gets wrong

It is the most repeated advice given to new mothers and, for a mother with insomnia specifically, it makes things worse in two ways.

It adds failed attempts. Lying down when you are not sleepy and not sleeping is practice at not sleeping, and the bed keeps the score.

And it spends the pressure. Sleep pressure builds across the hours you are awake and is the main thing you have working for you at 10pm. A ninety-minute afternoon sleep spends a large part of it.

Rest when the baby sleeps - lie down, put your feet up, do nothing. Do not try to sleep. That distinction sounds like hair-splitting and is the whole thing.

Take the free assessment to see:

✓Your Insomnia Severity Index (ISI) score
✓A personalized 6-week CBT-I plan, built by a sleep physician
✓How your sleep improves, night by night

Start free assessment

The protocol, modified

Standard sleep restriction assumes you control your nights. You do not. The published trial that tested CBT-I from pregnancy through six months postpartum handled this by modifying the protocol rather than dropping it, and it worked - CBT-I lowered insomnia severity across the postpartum timepoints with a medium effect, while a responsive bassinet, tested in the same trial as a comparison, did not.

Three modifications carry most of it.

Anchor the rising time, not the bedtime. The morning is the only end of the night you still control. A fixed wake time is what holds the body clock steady, and it is doing that job whether or not the night behind it went well. Bedtimes will be chaotic for a while. Let them be.

Count opportunity, not sleep. Your window is the stretch you are allowed to be in bed. Night feeds happen inside it. You are not trying to hit a total; you are trying to stop the window sprawling to eleven hours in pursuit of six.

Do not go below five and a half hours. In that trial, prescribed windows were never shorter than that and were set collaboratively rather than imposed. There is a floor in normal CBT-I too, and postpartum is the wrong time to go near it.

The part that is not a sleep problem

Some of what looks like postpartum insomnia is postpartum depression or postpartum anxiety, and a sleep program is the wrong tool for either.

Signals worth taking seriously: you feel wired rather than tired. You cannot sleep even when someone else is fully on duty and you have permission to. There are intrusive thoughts, or dread, or low mood that does not lift when you do get a good stretch. Or the anxiety is not really about sleep at all - sleep is just where you notice it.

None of those mean you cannot also have insomnia. They mean the order of operations changes, and someone should be looking at the whole picture. Bringing two weeks of diary to that appointment makes it a much more useful conversation than describing a bad month from memory.

It is worth saying plainly that professional guidance here is thin. ACOG's guidance on postpartum care mentions sleep only in passing - discuss coping options for fatigue and sleep disruption - and does not recommend CBT-I or anything else specific. That gap is real, and it is part of why this goes unaddressed.

What to expect

Slower than the six weeks a normal program runs, because your nights are genuinely being interrupted and you cannot control that variable. The rising time still holds. The window still stops sprawling. The bed still stops being a place you lie awake.

And when the baby does start sleeping through, you will not discover that you have quietly become someone who cannot - which is the actual thing this prevents.

More on sleep across a woman's life, including what changes through perimenopause.

Somnera is a self-guided education program built on CBT-I, founded and written by Dr. Camilo Ruiz, DO, FACOI, FAASM. It is not a diagnosis and does not replace care from your own clinician - particularly in the first year after a birth. The assessment is free and takes about two minutes.

Citations

Takenoshita M, et al. Insomnia disorder at 6 weeks postpartum: a prospective study. J Clin Sleep Med. 2026;22:152.
Quin N, et al. Preventing postpartum insomnia: findings from a three-arm randomized controlled trial. Sleep. 2024;47(8):zsae106. doi:10.1093/sleep/zsae106
ACOG Committee Opinion No. 736. Optimizing postpartum care. Obstet Gynecol. 2018;131:e140–e150.
Edinger JD, et al. Behavioral and psychological treatments for chronic insomnia disorder in adults: an American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med. 2021;17(2):255–262. doi:10.5664/jcsm.8986

Frequently asked questions

The test is what happens when you get an opportunity. Newborn sleep deprivation means you fall asleep fast whenever you can. Postpartum insomnia means the baby is asleep, the house is quiet, you are exhausted, and you still cannot drop off.

Not the standard version. The trial that tested CBT-I through pregnancy and the postpartum year modified it, and never set a window below five and a half hours. What gets anchored is the rising time, not the bedtime.

It is the most-repeated advice and it works badly for a mother with insomnia. Lying down when you are not sleepy adds failed attempts, and daytime sleep drains the pressure you need at night. Rest is fine. Trying to sleep is the part that backfires.

If you cannot sleep even when the baby is sleeping and someone else is on duty, if you feel wired rather than tired, or if there is low mood, intrusive thoughts, or anxiety that does not track with tiredness. Those point at postpartum depression or anxiety, which need a clinician, not a sleep program.

In a 2026 study that used structured diagnostic interviews at six weeks postpartum, 13% of women met full criteria for insomnia disorder, and a quarter had insomnia symptoms. Of those with symptoms, almost none had been clinically evaluated for it.

Perimenopause insomnia: when the hot flashes settle and the waking does not

Vasomotor symptoms start it. Something else keeps it going - which is why women on HRT so often say the flashes are better and the sleep is not.

Sep 2, 20265 min read

A staircase climbing in steps, with one long level landing marked off between two rules.

Sleep restriction

Sleep restriction isn't working. What to check before you quit

Week four, no change, and everyone else seems to have improved by now. Usually the protocol has not failed - it has not been titrated. The decision rules, in order.

Sep 2, 20265 min read