Co-sleeping: how long it lasts and how to end it gently

A far more common practice than people think

Co-sleeping is very common among preschool-age children. In a cohort of more than 900 children followed in pediatric care, 38% practiced reactive co-sleeping, meaning they joined the parents' bed partway through the night rather than sleeping there from bedtime. Nearly one child in four (23%) did so every night.

That figure drops with age: in the same study, nearly half of the children who joined the parents' bed every night had stopped two years later. For most families, co-sleeping is a stage, not a fixed habit.

Is co-sleeping dangerous for a baby?

Before age one, bed-sharing raises the risk of sudden infant death syndrome, and the American Academy of Pediatrics recommends avoiding it. The risk rises further with parental smoking, alcohol or drug use, extreme fatigue, or a soft sleep surface (a couch, a mattress that's too soft, pillows, a thick comforter). It is also higher before 4 months and for babies born prematurely or small. Pediatric guidelines favor a nearby but separate sleep space during the first months.

In France, the HAS recommends putting the baby in their own bed, ideally in the parents' room for the first six months, without sharing the parents' bed. French health insurance (ameli.fr) gives the same guidance. It adds that you should not keep the baby in your bed if you are exhausted or have taken medication that reduces alertness, and that you should not fall asleep with the baby in your arms.

Even without smoking or alcohol, the risk exists for a very young baby. An analysis of five case-control studies looked at breastfed babies under 3 months with no other risk factors, whose parents didn't smoke. These babies had a higher risk of sudden death in the parents' bed than in the parents' room on a separate surface. The absolute risk stayed very low in both cases, and the risk tied to bed-sharing fell as the baby got older.

Past the first year, the documented risk shifts in nature: a firm mattress, no gap where the child could get stuck between the bed and the wall, and age-appropriate bedding cover the essentials. When in doubt about a specific situation, a pediatrician's advice remains the reference, not an online guide.

Before age one, one distinction meaningfully changes things: sharing the room without sharing the bed. The baby sleeps in their own crib or bassinet placed near the parents' bed, which the American Academy of Pediatrics recommends for at least the first six months. This arrangement can cut the risk of sudden infant death syndrome by up to 50%, while remaining markedly safer than actually sharing the bed. The word "co-sleeping" therefore covers two practices with very different risk profiles: the precautions above are aimed at bed-sharing, not at a parent's presence in the same room.

Couches and armchairs are to be avoided in all circumstances with a baby: sleeping on one with an adult is the situation where the American Academy of Pediatrics describes the highest risk. If you worry you might fall asleep while feeding or comforting your baby in your bed, remove pillows, sheets, and blankets that could cover their face. Put them back in their own sleep space as soon as you wake up.

How long can a child co-sleep?

After age one, there's no official age limit for co-sleeping: professional bodies no longer set a recommendation on shared sleep, for lack of data clear-cut enough to draw a rule from. In practice, co-sleeping stays common through preschool, then drops off noticeably once children start elementary school.

The trajectory isn't linear either: a child who's been sleeping alone for months might ask to come back to the parents' bed after an illness, a move, or the arrival of a younger sibling, without that return signaling a lasting problem. These temporary comebacks generally resolve on their own once the triggering event has passed, following the same gradual pattern as the original transition out.

Why families end up there

A family rarely arrives at co-sleeping for just one reason. Nighttime breastfeeding often finds natural ground there: feeding without having to get up makes the first months' nights easier, for mother and child alike. Other families arrive at it by cultural or family choice, since a shared bed is the norm, not the exception, across much of the world. Others end up there after one hard night: an isolated waking turns into a habit, simply because everyone sleeps better that way once it's settled in.

For the child, asking to join the parents' bed can also reflect an anxiety specific to bedtime rather than a preference for comfort. Separation anxiety drives some children to demand a presence they wouldn't accept losing, even in a very comfortable bed. Our guide on separation anxiety covers this mechanism and its age peaks in detail. It's worth a look if the refusal to sleep alone comes with crying at the slightest departure, rather than a simple rejection of the bed itself.

Ending co-sleeping gently

To end co-sleeping, a gradual transition works better than switching overnight. You start with a spare bed or a mattress on the floor in the parents' room, then move that fallback spot toward the child's own room in stages. The child gets time to associate their own bed with a safe place rather than with being left alone.

Announcing the change ahead of time and keeping the bedtime ritual identical throughout the transition limit resistance. It's also better to pick a calm moment, one that doesn't land alongside a first day of school, a move, a new sibling, or the end of the nap. Novelty goes down easier when everything else stays put.

Setting up the move to independent sleep, step by step

To set up the move to independent sleep, a few benchmarks help build the plan. With a child old enough to understand the change, involving them in the decision turns it into a project rather than an imposed constraint: picking their own bedding, decorating their corner of the room, naming a stuffed "guardian" that stays with them at night. A simple tracking chart, a sticker or a drawing for each successful night, works well from about 3-4 years old. The reward doesn't need to be material: the pride of checking the box is often enough.

An older child, aware of the change, tends to negotiate more than a toddler who's simply moved into a new bed: promises of compromise, haggling over bedtime, repeated requests to stay "just this once." Setting the rule once, clearly, and sticking to it without reopening the discussion every night keeps bedtime from turning into a permanent negotiation. Consistency in the framework matters more here than how strict it is.

Gradually withdrawing the parent's presence pairs well with the step-by-step move of the sleep spot already described above. You sit near the bed for the first nights, then move the chair a little closer to the door every two or three nights, until you're fully out of the room. Each step should only be taken once the previous one has settled: it's better to stretch out a stage than to go back once it hasn't worked.

With several children who'd been sharing the same parents' bed, transitioning them all at once needlessly complicates things. Having them sleep together in the same room, at least at first, leaves them a familiar presence while the parent's disappears: the sibling takes over as the landmark instead of losing both at the same time. If the same period also calls for a switch from a crib to a bed, it's better to keep the two projects separate. Sort out the bed first while keeping the sleep location unchanged, then tackle independent sleep afterward. Our guide on the transition from crib to bed covers that step in detail. It's a project in its own right, which deserves its own attention rather than being piled on top of this one.

If a waking brings the child back to your bed

A child who comes back once or twice a week after a nightmare, a teething spell, or a sleep regression isn't undoing the transition. Walking them calmly back to their own bed, without a long discussion or negotiation, is usually enough to keep these returns occasional.

What matters is the trend over several weeks, not the one-off night when tiredness got the better of the best intentions.

A surge in wakings lasting more than two or three weeks deserves a different read than an isolated return. It more often points to an identified sleep regression, a lingering illness, or a passing stress (starting school, a move, a new sibling) than to a genuine step backward on the transition itself. Continuing to walk the child back without reopening the discussion, while looking for the likely cause during the day, avoids mistaking a passing episode for a failure of the new sleep setup.