Sleep regressions: the real timeline by month

"Sleep regression" has become a catch-all term for any week where your baby's sleep got worse. Some regressions are real, well-documented phases tied to specific developmental milestones. Some are made up by Instagram. Here is what the evidence supports, month by month.

4 months: REAL (and permanent)

The big one. Sleep architecture matures from 2-stage to 4-stage adult-style sleep. This is not a phase. It is a one-way change. Once it happens, the patterns of falling asleep and night waking shift permanently. I wrote about this in detail in The 4-month regression isn't a regression. The short version: it is real, it does not pass on its own, and the window of opportunity to teach independent sleep is wide here.

6 months: NOT really a thing

If your baby's sleep got worse around 6 months, the actual cause is more likely solids introduction (tummy adjustment), the 3-to-2 nap transition starting, or teething. There is no documented neurological "6-month regression." The term shows up on parenting blogs but not in the developmental literature.

8 to 10 months: REAL

This is one of the most disruptive. Three things happen at once. Object permanence emerges, which means your baby now knows you exist when you leave the room and can ask for you. Separation anxiety appears for the first time. Major motor milestones (crawling, pulling up, sometimes early walking) demand practice, and babies will practice in the cot at 11pm.

What to do: give massive amounts of floor time during the day for movement practice, hold the boundary on the routine at night, expect 2 to 3 weeks of disruption. This regression genuinely passes.

12 months: NOT really a regression, but a nap transition

What is labelled the "12 month regression" is almost always the 2-to-1 nap transition starting. This shows up as nap refusal, late morning meltdowns, and short afternoon naps. It is not a sleep regression. It is a daytime schedule issue.

Hold off on the 1-nap transition until 13 to 16 months for most babies. Cut the morning nap shorter rather than dropping it entirely first. The transition is gradual for most babies and complete by around 15 months.

15 months: marginal

The literature here is thin. If your baby's sleep is disrupted at exactly 15 months, the most likely cause is the 2-to-1 nap transition (if not already complete), a language burst, or molars. Not a discrete regression.

18 months: REAL-ISH

The mechanism is not as cleanly documented as 4 or 8 months. Two real factors play in. Independence and a developing sense of self: toddlers start to push against being put to bed. Language explosion: cognitive activity at night increases temporarily. Add in molars and you have a meaningful disruption window.

Hold the line on bedtime and the routine. This window typically lasts 2 to 4 weeks.

24 months: REAL, but as new types of waking

Around 2 years old, nightmares can begin. Toddler imagination matures enough to generate scary content. Night terrors (a different phenomenon, occurring in deep sleep) can also start.

Nightmares respond to morning conversation about the dream and a comforting bedtime routine. Night terrors do not. Night terrors look terrifying but the child is not awake; do not engage, do not wake them, ensure safety, wait it out. If terrors are frequent, sometimes a "scheduled awakening" 30 minutes before the typical terror time disrupts the pattern.

The two myth regressions to ignore

The 11-month regression. Not a thing in the literature. If you see this in your child, look at the 2-to-1 nap transition window starting earlier than average.

The 3-year regression. Not documented. If sleep deteriorates around 3, look at the nap dropping for most children, a new sibling, a new bedroom, or a new daycare environment as the more likely cause.