Quick answer: Night waking is a physiological norm in the first year. The strategies making the most difference: split nights so each parent gets one consolidated sleep block; optimise the sleep environment; and from 4–6 months, consider sleep training if the situation is unsustainable.
Reframing what’s happening
Human sleep cycles approximately every 90 minutes, transitioning from deep sleep toward lighter sleep. Adults manage this transition automatically and don’t recall brief awakenings. Babies who haven’t learned to link sleep cycles fully wake at the transition, require settling, and need to re-establish the conditions under which they originally fell asleep. If those conditions were feeding or rocking, those conditions are needed at 2am. This is physiologically normal and not a behaviour problem. Understanding this reframes the objective: the goal isn’t fixing a broken baby but supporting a developing skill (independent settling).
The split-night strategy
The most evidence-supported approach for managing parental sleep deprivation: split the night so each partner gets one consolidated sleep block. One partner takes all wakings from bedtime to 2–3am; the other takes everything from 2–3am to morning. Each gets approximately 4–5 consecutive hours. Research on sleep deprivation is consistent: 4–5 consecutive hours produces significantly better cognitive function than the same total hours in fragments. Even one consolidated block per night changes the subjective experience of parenthood substantially.
Environment: the highest-return investment
Blackout blinds: light entering at the 45-minute sleep cycle transition wakes babies who would otherwise link to the next cycle. This one change can extend naps and reduce early morning waking significantly. White noise running all night: masks ambient sounds causing arousal at cycle transitions. Room temperature 16–20°C: overheating is associated with more disturbed sleep and SIDS risk. These are set-once changes with ongoing benefit — disproportionately valuable relative to the effort.
Sleep training from 4–6 months
From 4–6 months, developmental readiness for sleep training is established. Graduated extinction (Ferber), full extinction, and gentler fading methods all have evidence for effectiveness. Long-term follow-up studies (including a 5-year follow-up by Price et al.) find no adverse effects on attachment, wellbeing, or child behaviour from sleep training. The decision is personal — ‘wait it out’ and ‘sleep train’ are both valid. The key variable: is the current situation sustainable? Unsustainable parental sleep deprivation affecting mental health and safety warrants addressing.
The science behind the split-night approach
Sleep deprivation research consistently distinguishes between total sleep hours and sleep architecture. The critical variable for cognitive function, reaction time, and emotional regulation is not total hours but the length of the longest continuous sleep episode. Four hours of consecutive sleep produces significantly better next-day functioning than eight hours in 90-minute fragments — even though the total is lower. This is why the split-night approach (each parent taking a defined block) is more restorative than both parents responding to every waking. The parent sleeping from 10pm to 3am gets a full sleep cycle and a half; the parent sleeping from 3am to 8am gets the same. Both have had a functional sleep block even if total hours remain low. The same mathematics applies to a single parent using a night nanny, a trusted family member, or any other arrangement that produces one consolidated sleep episode.
Frequently Asked Questions
Is it safe to drive on broken sleep?
Research shows chronic partial sleep deprivation produces driving impairment equivalent to alcohol. After 17–19 hours awake, impairment is equivalent to 0.05% blood alcohol. Don’t drive if you’ve had less than 4 hours sleep in the previous 24 hours or have been awake more than 17 hours. This is evidence-based safety guidance, not overcaution.
My baby was sleeping through and suddenly started waking every 2 hours — why?
Sleep regressions — periods of disrupted sleep following consolidation — are associated with developmental leaps and occur at approximately 4 months, 6 months, 8–10 months, and 12 months. They correlate with major developmental advances (motor development, object permanence, language). Most regressions last 2–6 weeks before sleep returns to the previous pattern.
When does night waking improve without sleep training?
For most families who don’t sleep train, significant improvement comes between 6–12 months as sleep architecture matures and the baby becomes more able to self-settle at cycle transitions. Most babies without sleep training are significantly better by 12–18 months.
Related Reading
- Sleep deprivation as a new parent: science-backed coping strategies
- Baby sleep schedule generator by age
Found this helpful? Sign up to the LylyMama newsletter — evidence-based answers to the questions every new parent actually has.
