Does a baby's flat head go away on its own?

Most flat spots follow the same arc. Parents typically first notice flattening at around six to eight weeks, because that is roughly how long it takes for a flat area to form. Severity peaks at about four months. New flattening after three months is unusual, because by then most babies can hold their head up against gravity and move it themselves. From four months onward rolling, sitting and crawling take the pressure off the back of the head, and the flattening stops getting worse. What it does not do is fill back in. Nothing pushes the flat area out again, so the head grows on around it and the flattening becomes a smaller share of a bigger head.

Those timings come from Dr Gary Rogers, a craniofacial surgeon, in his 2011 review of deformational plagiocephaly in the Journal of Craniofacial Surgery. His paper is the reference point for this article, and the studies he draws on are cited at the end.

If your baby is ten weeks old and the flat spot looks worse than it did a fortnight ago, you are not watching something go wrong. You are watching the normal pattern, at the normal time.

Not in the way parents are usually told. A flat spot forms because steady pressure on one part of the skull holds it back while the rest of the head keeps growing. Once your baby has the head control to move themselves, that pressure comes off and the flattening stops getting worse. What it does not do is fill back in. Nothing pushes the flat area out again, so the head grows on around it and the flattening becomes a smaller share of a bigger head. That is why measured rates fall so steeply while more than half of parents still notice the shape years later. What nobody can predict is how much of it your baby's own growth will disguise.

Because flattening comes down to steady pressure on one part of the skull, what your baby rests on during those long sleep stretches matters. The Perfect Noggin is a contoured sleep surface that redistributes pressure away from flat spots and makes varying the point of pressure much easier for a baby with strong side preference.

The usual explanation is that a baby's skull is soft, so it squashes. Dr Rogers dismisses that. As he points out, if you lay a newborn on a firm flat surface, the head does not deform. And if softness alone were the cause, every back-sleeping baby would end up flat. In practice only a minority do. In the Hutchison cohort, around one in five infants had measurable flattening at four months.

What actually happens is about growth, not squashing. The skull does not push itself outward. It grows passively in response to the brain expanding underneath it, fastest in early infancy and tapering sharply after the first year. Where the head rests against a mattress, that growth is resisted, so the extra volume goes elsewhere. The result is a flat area with compensatory bulging around it.

Rogers uses a pumpkin to explain it. A pumpkin in a field cannot grow into the ground, so it grows along it and flattens underneath. A bigger, faster-growing pumpkin presses down harder and flattens more in the same time.

That explains why flattening takes six to eight weeks to appear rather than showing up overnight, and, Rogers suggests, why boys are affected more often, since they have larger and faster-growing heads on average.

Rogers is direct about this: the babies who flatten are the ones who cannot easily change their own head position in the first few months. The single biggest reason is a cervical muscle imbalance, a neck that does not turn evenly in both directions. Fenton and Gaetani report cervical imbalance in 70 to 95 per cent of infants with deformational plagiocephaly, a figure Rogers also reports in his 2011 review. In some babies this is congenital muscular torticollis, which only a clinician can diagnose. It develops when the head is held rotated and tilted in the womb long enough for one sternocleidomastoid muscle to shorten and the opposite one to lengthen and weaken.

That explains something parents notice and rarely connect. Rogers explains that most babies engage the pelvis with the head turned to the right, so affected babies tend to lie on their right occiput, and he notes nearly every major study has found more right-sided flattening.

The earliest sign is not a visible head tilt. It is a head position preference: your baby keeps returning to the same side however you reposition them. Rogers' view is that a preferred head position should be treated as a possible cervical imbalance until a clinician has checked it. If that sounds like your baby, it is worth raising with your child and family health nurse, GP or paediatric physiotherapist, who can assess neck rotation properly.

Rogers lists what else raises risk and why. Prematurity and developmental delay, because both delay independent head movement. Twins and other multiples, first-born babies, assisted deliveries and breech presentations, because all four increase the chance of the head being constrained in the womb.

Not on its own, no. The American Academy of Pediatrics recommended back sleeping in 1992, and the Back to Sleep campaign followed in 1994. By 2000 the AAP was reporting a decline of more than 40 per cent in sudden infant death syndrome in the United States. Cases of flattening rose sharply over the same period. That correlation is real.

But Rogers is clear that back sleeping cannot be the cause on its own. Flattening also occurs in babies placed on their front, and was described long before back sleeping became standard practice in the West. Most babies placed on their back never develop clinically significant flattening. His conclusion is that back sleeping amplifies other risk factors rather than causing the problem.

Practically, back sleeping is not the thing to change. It is a cornerstone of safe sleep guidance in Australia and should stay exactly as it is. Follow the current Red Nose Australia safe sleep recommendations. What you change is everything around it: tummy time, how you carry and feed, and getting the neck assessed if there is a position preference.

Flattening becomes visible at around six to eight weeks and peaks at four months. Measured rates fall from there. The prevalence figures below are from Hutchison and colleagues, a prospective cohort of 200 New Zealand infants followed from birth to two years, and they are the study Rogers relies on for the natural history. Read them as the share of babies scoring above a measurement threshold at each age, not as flat spots that have gone. Every baby in the cohort was receiving ordinary care, so there is no untreated group to compare against.

• Birth to 6 weeks. Birth moulding settles. The most important window for prevention
• 6 to 8 weeks. Flattening becomes visible. This is when most parents notice
• 3 months. Most babies can hold their head against gravity. New flattening after this is unusual
• 4 months. Severity peaks. Plagiocephaly present in 19.7 per cent of the Hutchison cohort
• 4 to 6 months. Rolling begins. Any head tilt tends to show itself now
• 8 months. Sitting takes the pressure off the back of the head. 9.2 per cent of the cohort
• 12 months. Skull growth has tapered sharply. 6.8 per cent of the cohort
• 2 years. 3.3 per cent of the cohort still above the threshold. Residual asymmetry below that threshold is common and does not show in this figure

Plenty of babies arrive with an odd-looking head from the birth canal, or from a vacuum or forceps delivery. Pregnancy, Birth and Baby notes a baby's head generally returns to its normal shape within about six weeks of birth.

This is also the most important window for prevention. The NSW Health guideline for allied health professionals, GL2020_013, identifies the first six weeks as the most important period for preventing head shape deformity. Alternating which end of the cot you use, alternating which arm you feed from, and supervised tummy time from birth count for more here than at any later point.

This is when most parents first see something, and it matches the mechanism exactly. Rogers puts the average at six to eight weeks, because that is how long sustained contact takes to produce visible flattening. Raising Children Network suggests having a head shape checked if a flat spot has not gone back to a typical shape by about two months, and if there is also a head position preference, raise that at the same appointment. If you are not sure what you are looking at, the four-point check walks through it in two minutes.

Severity then peaks at around four months. Two things converge. The head is at its heaviest relative to your baby's ability to move it, so pressure is greatest, and most term babies can support the head against gravity by around three months, after which Rogers notes further flattening is unusual. So the peak is not the flattening still deepening. It is the accumulated effect becoming most visible just as the cause runs out of runway.

Rolling changes things. Raising Children Network puts rolling both ways at around four to five months, and a baby who rolls in their sleep stops resting on one spot. Once your baby is rolling, keep placing them on their back to sleep and let them find their own position, in line with Red Nose Australia guidance.

This is also when a head tilt often appears for the first time. Before about three to four months the head is supported by a surface most of the time, so the neck muscles are never really tested. Once your baby holds their own head up, any imbalance shows. Rogers notes a tilt at five or six months is often weakness on the opposite side rather than tightness on the side of the tilt.

That distinction is exactly why this is a job for a paediatric physiotherapist rather than something to work out at home. If a flat spot has not started to improve at all by around four months, that is the point to get it assessed rather than give it more time. The 3 to 6 month window covers what is realistic from here.

Sitting without help typically develops somewhere between five and seven months depending on the source. Pregnancy, Birth and Baby states flat head usually improves once a baby can sit up, and the cohort figures point the same way: Hutchison fell from 19.7 per cent at four months to 9.2 per cent at eight months. It is worth knowing what that figure is, though. It is the share of babies scoring above a measurement threshold, and because the measurements are ratios, a head that keeps growing scores better even when the flat area itself has not changed.

If flattening is severe and has not improved despite repositioning and physiotherapy, this is the age range where a helmet gets considered. The Royal Children's Hospital notes helmet therapy is most effective if started between six and eight months of age, and Pregnancy, Birth and Baby notes fewer than one in ten babies with plagiocephaly will need one.

Measured rates keep falling, more slowly. Hutchison recorded 6.8 per cent at twelve months and 3.3 per cent at two years. Your baby is crawling and pulling up, so there is almost no pressure on the back of the head while awake. Some of that fall is the head growing rather than the flat area filling in, because the measurements are ratios and a fixed flat area becomes a smaller share of a bigger head.

A Dutch cohort study by van Vlimmeren and colleagues followed children to five years and found head shape within the normal range for 80 per cent, mild in 19 per cent and moderate or severe in 1 per cent. Longer follow-up is less tidy. Steinbok followed children to beyond five years of age and residual asymmetry was still noticed by 58 per cent of parents, though only 21 per cent were concerned about how it looked. Hair covers a good deal of what remains, which is part of why the older figures read better than the younger ones.

Find your baby's age on the timeline and read the two stages either side of it. In the first four months you are in the part of the curve where what you do has most effect. Past six months, change is slower and harder to see, and nobody can tell you in advance which group your baby is in. That is the argument for getting the neck assessed now rather than giving it another few months.

The most useful step is the same at every stage, and it follows from Rogers' argument. Because most flattening traces back to a neck that will not turn evenly, book with your child and family health nurse or GP and ask for a referral to a paediatric physiotherapist. A physio can measure the asymmetry, test neck rotation properly, and work out whether the issue is tightness, weakness or both.

Photograph your baby's head from directly above every two weeks, in the same spot and the same light.

If your baby is under three months, the maths is different.

Everything above describes what usually happens to a flat spot over time. What it does not tell you is what to do in the weeks where it is still forming. Rogers notes further flattening after four months is unusual, and NSW Health identifies the first six weeks as the most important time for preventing head shape deformity. Before four months, you are not correcting something that has settled. You are in the window where it is still accumulating, and where the overnight hours do most of the work.

That is why the early months are worth acting in rather than waiting through. Book a paediatric physiotherapy assessment, keep the positioning and tummy time going, and if you are weighing whether The Perfect Noggin has a place alongside those, our suitability guide walks through the four things that decide it: your baby's age, whether rolling has started, how pronounced the flattening is, and whether the neck is involved.