Get access to over 1,300 articles and reader Q&As.
As with many experiences in parenting, just when you think you’re getting the hang of something, you’re thrown for a loop. Sleep (or lack thereof) is one of the more common aspects of parenting that feels out of your control.
When people talk about “sleep regressions,” they’re talking about any disruption to their baby’s sleep schedule (waking more at night, having trouble staying asleep for naps, or resisting bedtime). Typically, parents report that they felt like things were going well and then, all of a sudden, they were not.
Sleep regression concerns come up more at some times than others. Many new parents hear about the “4-month sleep regression,” a time in which (theoretically) babies all of a sudden stop sleeping much, and you become even more exhausted. “How long does the 4-month sleep regression last?” is one of our top Petey questions.
There is a more basic question, though: “Is the 4-month sleep regression a real thing?” More generally, are any sleep regressions real? Is this whole thing just made up? And whether it is real or not — how can I get my kid back to sleep right now?
Let’s unpack!
How does infant sleep vary over time?
Over the first year of life, sleep consolidates, first at night and then during the day. If we step back to the broader picture over the first year, wake-ups go from many times a night to an average of less than one wake-up per night once your baby turns 1. In the survey data in the articles linked above, almost no babies sleep through the night in the 0 to 3 month range, but about 60% of them do by the 10-12 month age range.
This is an important note, especially for new parents: Even if it does not seem like it right now, you should expect sleep to get better over time.
While this is the broad trend, within this, there is a lot of variation both across babies and also within a baby over time. A study using detailed infant tracking demonstrated that nighttime sleep can easily vary by two hours on a day-to-day basis. So, while your baby is sleeping more on average over time, today is not guaranteed to be better than yesterday. If you looked at a graph of hours of sleep relative to age, you would see an upward slope, but with a lot of variability.
The data behind the “4-month sleep regression”
Many new parents are warned of the dreaded “4-month sleep regression.” In popular discussion, the 4-month sleep regression is a time when babies all of a sudden wake up more frequently. In the data, we would expect to see this manifest in shorter sleep periods and more night wakings at this age range.
That pattern doesn’t appear starkly in the data. A 2016 paper using data from an iPhone app on sleep patterns shows an increase in nighttime sleep session length through the first year, with no dip at 4 months. A giant review study from 2018 argues that, on average, night wakings decrease over the first year. They note some parent-reported increases in night wakings from 3 to 6 months, but this is not consistent with wearable-based data. (Basically, parents may say there are more wake-ups, but when you detect them with technology, you don’t see that.)
A small (60 women) but detailed study used both parental reports and technology-based measures to look at sleep duration by week of life. The longest sleep period increased from 0 to 18 weeks, largely consistently across this time frame. There isn’t any obvious decrease in sleep duration in the 4-month range.
So, are sleep regressions real?
A sleep regression isn’t a formal term, and it really just refers to any situation in which sleep gets worse and you do not expect it. The reality is that this happens a lot on a day-to-day basis and isn’t consistent across babies once they hit a certain age. In a sense, every day that sleep is worse than the day before is a regression. But — glass half full — every day that sleep is better than the day before, it’s a promotion!
Unfortunately, the way new parenting often works, you’re more attuned to when things get harder than to when they get easier.
Based on the overall trends and the variability in the data, I think it is fair to say that “sleep regressions” are just a poorly branded term for normal sleep variability.
So if these sleep disruptions can simply be blamed on normal variability, why do perceived “regressions” seem to last longer than just a day or two?
A primary reason is our reaction to them. There is a cardinal rule in developing good sleep patterns in babies and children. It is: be consistent. Consistent bedtime routine and consistent approach to when you respond and when you do not. If your baby wasn’t crying before, and now they are, this is a time to develop that routine and plan. Babies form habits quickly, both good and bad (this is why sleep training often only takes a day or two).
If a perceived sleep regression changes how often you respond at night, or how you respond, that habit may persist. If you start co-sleeping in response to a couple of tough nights, you may struggle to go back to your previous sleeping arrangement. That might be okay with you! But often parents tell me these short-term sleep regressions turn into long-term experiences they do not like.
How can I improve my baby’s sleep?
Regardless of the reason for disrupted sleep, many parents are looking to improve it. If that’s you, I have two notes.
First, if you have a night or two of disrupted sleep, try not to change whatever routines you have. Do not assume this is a new normal! It’s very likely just normal, day-to-day variability. Stay the course, whatever it is. You will get through it. (This advice is based on virtually all sleep programs.)
Second, although sleep training is not for everyone, it is an option for many people and does tend to improve sleep for both kids and parents. If that direction is of interest, I’ve written before about whether sleep training is bad (it isn’t), and what the best sleep-training method is (one you can stick to).
The bottom line
- Infant sleep generally consolidates over the first year, moving from frequent night wakings to longer stretches and more consistent sleep by 12 months. Overall, this is the general pattern, but babies’ sleep varies a lot — both from one baby to another and for the same baby day-to-day.
- Many new parents are warned of the dreaded “4-month sleep regression,” but research doesn’t show a clear drop in sleep at this stage. Although parents report noticing more wake-ups, the data doesn’t support this.
- The reality is that sleep variability is common and happens a lot on a day-to-day basis — it isn’t consistent across babies once they hit a certain age.
- Perceived “sleep regressions” often last because parents change their responses — like starting new routines or habits — that babies quickly adapt to and maintain.
- If you’re hoping to improve your baby’s sleep, the key advice is to stay consistent through short disruptions and consider sleep training if it feels right for your family.
Log in
It’s interesting that this piece doesn’t look at the data factor that many online sleep people cite as being the reason for the 4 month sleep regression – a fundamental change in the way your baby sleeps around that age when they all of a sudden start being aware that they are waking after every sleep cycle. I have a 9 month old and she went through a wild, demonstrable change to her sleep right around the 4 month mark that resulted in her waking up every 45 mins (42 minutes really – that was the length of her average sleep cycle) overnight for weeks. We kept everything the same with her bedtime routine, and the waking only ended up being solved through CIO sleep training. I’d be really interested to see if there was any data to support a biological change around the 4 month mark because it just seems so odd that it would just be coincidental that so many babies would start to sleep more poorly than they had been before in that window.