Key Takeaways

Every age on the CDC checklist is the point by which 75% of children can do the skill, not the average. One child in four is later by design ¹².

The CDC says its own milestones are “not developmental guidelines or standards” and should not be used as screening or diagnostic tools ¹.

Motor ranges are far wider than any chart suggests. In a five-country WHO study, walking alone spanned 8.2 to 17.6 months ³.

4.3% of healthy children never crawled on hands and knees, which is one reason crawling is not treated as a required step ³.

A missed milestone is a reason to screen, not a reason to wait. The 2022 revision was made specifically to shorten “wait and see” ¹⁴.

The short answer

A milestone chart answers one question: by this age, can most children do this? It does not answer whether your child is developing normally, and it was never designed to.

Age	What most babies can do by then ⁵
2 months	Calms when spoken to or picked up; looks at your face; smiles when you talk or smile
4 months	Smiles on their own to get your attention; chuckles; makes cooing sounds; turns toward your voice
6 months	Knows familiar people; laughs; takes turns making sounds with you; reaches to grab a toy
9 months	Shy or clingy with strangers; looks when you call their name; makes lots of different sounds; lifts arms to be picked up
1 year	Waves bye-bye; calls a parent “mama” or “dada”; understands “no”; puts a block in a cup; pulls up to stand
15 months	Copies other children; shows you an object; claps when excited; tries to say one or two words besides “mama” and “dada”
18 months	Points to show you something interesting; tries to say three or more words; follows a one-step direction without gestures

Read the column header again. It says most, and the CDC defines most as 75% or more ¹. That one word is the difference between a checklist and a test.

Where the month on the chart comes from

The current CDC checklists were published in February 2022, replacing a set that had been in use since 2004 ¹⁴. The revision was not cosmetic.

The older checklists placed milestones at something close to the average age, which is roughly the 50th percentile. Half of children would reach the skill before that date and half after, so a child who had not done it was statistically unremarkable. The working group convened by the AAP changed the anchor: each milestone now sits at the age by which at least 75% of children would be expected to do it ²⁴. The revision also added checklists at 15 and 30 months, so that every recommended well-child visit from 2 months to 5 years has one ⁴.

The CDC explains why in one sentence worth quoting. With a milestone at 50%, “half of all children would still reasonably be expected to have not yet achieved it,” and that makes a missing milestone easy to explain away ¹. At 75%, a missing milestone is unusual enough to act on. The stated goal was to reduce “the ‘wait and see’ approach often taken when a child is missing an average age milestone” ¹.

So the number is not a deadline, and it is not an average. It is a trigger for a conversation, set deliberately early enough that a child who is behind gets noticed while help still changes the outcome.

What the CDC says about its own checklist

This part surprises most parents who read the checklist carefully.

“The developmental milestones included on CDC’s milestone checklists are not developmental guidelines or standards for children’s development,” the agency states. “They should not be used as screening or diagnostic tools to detect developmental delays” ¹.

That is not a disclaimer buried in a footer. It is the design. The checklists are described as communication tools, built to start a conversation at a well-child visit and to flag when a formal screen is needed ¹. A checklist item is a prompt, not a verdict.

There is a practical consequence. If your child has not done something on the list, the answer is not to find a stricter chart. The answer is the next step the CDC itself names: screening with a validated tool, or a referral ¹⁴.

How wide normal really is

For motor skills there is better data than any checklist, and it makes the point more sharply.

The WHO Multicentre Growth Reference Study followed 816 children in Ghana, India, Norway, Oman and the United States, assessed monthly in the first year and every two months in the second ³. It reported windows of achievement bounded by the 1st and 99th percentiles, with no internal cut-offs.

Milestone	1st percentile	99th percentile	Span
Sitting without support	3.8 months	9.2 months	5.4 months
Standing with assistance	4.8 months	11.4 months	6.6 months
Hands-and-knees crawling	5.2 months	13.5 months	8.3 months
Walking with assistance	5.9 months	13.7 months	7.8 months
Standing alone	6.9 months	16.9 months	10.0 months
Walking alone	8.2 months	17.6 months	9.4 months

A child who walks at 9 months and a child who walks at 17 months are both inside the range that contains 98% of healthy children ³. That is a spread of more than nine months on a skill most parents treat as a fixed appointment.

Two more findings from the same study matter. About 90% of children achieved five of the six milestones in the same order, and 4.3% never crawled on hands and knees at all ³. Crawling is common, useful and optional.

The authors are careful about how the windows should be used: for descriptive comparison between populations, and “to signal the need for appropriate screening when individual children appear to be late” ³. They are not a standard to measure a child against, which is the same position the CDC takes about its own checklists.

One caution about reading these numbers beside a checklist: the studies define skills differently. WHO’s “sitting without support” is a tested position held for a set time, not a parent’s impression at playgroup. Comparing a checklist age with a percentile window compares two different instruments, and small differences between them mean nothing.

The milestones parents worry about most

Rolling over. This is the first motor skill with a visible date and the least standardized one. It is not among the six milestones in the WHO study at all, which tells you how much weight to give it. Some babies roll back to front first, some front to back, some take months to manage both directions, and some skip rolling and move straight to sitting and scooting. The thing to watch is not the roll but whether the baby is getting stronger and more mobile across weeks. Rolling also changes where the head rests, which is usually the point where a flat spot starts to improve; the flat head guide covers that timeline.

Sitting. The WHO window runs from 3.8 to 9.2 months ³. That early edge explains a lot of unnecessary comparison. A 4-month-old who sits briefly with support is at the front of a normal distribution, not ahead of schedule. A 7-month-old who still tips sideways is inside it.

Crawling. Wide window, and optional. A baby who bottom-shuffles, rolls across the room, or commando-crawls on their forearms is solving the same problem by a different route.

Walking. 8.2 to 17.6 months ³. A family worried at 14 months is usually worrying about a skill whose normal range extends three and a half months beyond their child’s age. The families who should ask are the ones past the window, not the ones near the middle of it.

First words and pointing. These come from the CDC checklists rather than the WHO study, and they carry more diagnostic weight than motor timing. Early language and joint attention are the skills that predict later communication. At 18 months the checklist looks for three or more words besides “mama” and “dada”, and for pointing to show you something interesting ⁵. Pointing to share attention, rather than to demand an object, is the one to notice.

Why “late” is usually not “delayed”

Three things explain most of the gap between a chart and a child.

The range is real. A nine-month spread on walking is not measurement error or a failure of the study. It is the distribution ³.

Prematurity shifts the whole calendar. If a baby was born early, developmental age is calculated from the due date rather than the birth date until age 2 ⁶. A baby born six weeks early is being measured against a calendar that runs six weeks ahead of their development, and a chart that does not correct for it will manufacture a delay that does not exist.

Skills cluster and stall. Children often plateau for weeks and then gain two skills in a fortnight. One visit is a snapshot. The trend across two or three months is the signal.

The short list that needs action

The point of the 75% rule is that a missing milestone should lead somewhere. These are the ones not to sit on.

What you notice	Why it is different	What to do
No social smile by 2 months	Smiling back is on the 2-month list ⁵	Raise it at the visit; do not wait for the next scheduled check
Does not look when you call their name by 9 months	Name response is on the 9-month list ⁵	Ask for a hearing test and a developmental screen
No babbling or varied sounds by 9 months	Sound play is the precursor to words ⁵	Ask for a screen rather than a second opinion from the internet
Not sitting without support by 9 months	The WHO window closes at 9.2 months ³	Book an assessment
No pointing and fewer than three words by 18 months	Both are 18-month items ⁵	Ask for a developmental screen and an autism-specific screen
Not walking alone by 18 months	The WHO window closes at 17.6 months ³	Book an assessment
Any skill that appeared and then disappeared	Loss is a different event from delay	Seek prompt review, not a wait-and-see period

That last row is the one to remember. A skill that has not arrived is a question about timing. A skill that arrived and then went away is a question about something else, and it does not belong in a wait-and-see category.

What screening actually involves

A screen is not a diagnosis, and it is not the checklist. It is a validated questionnaire or structured assessment with known sensitivity and specificity, scored against a population.

In the United States, the AAP recommends developmental surveillance at every well-child visit and standardized screening at the 9-, 18- and 30-month visits ⁷. Autism-specific screening is recommended at 18 and 24 months ⁸. In the United Kingdom, the health visiting team sends the Ages and Stages Questionnaire, ASQ-3, before the 9 to 12-month and 2-year reviews ⁶.

	United States	United Kingdom
Ongoing surveillance	Every well-child visit ⁷	Health visitor reviews until about age 2 ⁶
Standardized screen	9, 18 and 30 months ⁷	ASQ-3 before the 9–12 month and 2-year reviews ⁶
If a concern is raised	Screening, then referral to early intervention ⁷	Referral through the health visitor or GP ⁶

The distinction that matters: a checklist is filled in and interpreted alone, while a screen is scored against a validated cut-off. Confusing the two is how a missed milestone becomes either a false alarm or a missed referral, depending on the household.

Do milestones predict intelligence?

No, and the question confuses two different measurements.

A milestone records whether a skill has appeared by an age. An intelligence test measures performance on a set of tasks at a point in time, standardized against peers. The first is a developmental checkpoint. The second is a score. A child who walks at 10 months is not smarter than a child who walks at 15; they have different motor maturation, and both are inside the range ³.

What milestone timing does predict, weakly and mostly at the extremes, is the chance of a developmental condition worth assessing. That is what the checklists are built to flag ¹². Using them as an early intelligence test is a category error, and it makes parents anxious about the wrong thing.

Recording milestones without turning them into a scoreboard

The useful record is small: the date a skill first appeared, and a note of what it looked like. Not a score out of a checklist, and not a comparison with a cousin two weeks older.

There are two reasons. The date is what a clinician asks for, and it is impossible to reconstruct from memory at an 18-month visit. In the UK, the red book has a milestones section built for exactly this ⁶. And a record of firsts is a different object from a record of deficits. One is a keepsake that happens to be clinically useful. The other turns a wide normal range into a source of daily alarm.

Our milestone tracker logs a first with a date and a photo, and keeps it beside feeding and sleep, so the timeline is there when someone asks. If the logging itself has become the stressor, tracking without data anxiety is the counterweight. For the growth side of the same visit, the guide to growth chart percentiles explains why a drop from the 50th to the 25th is usually not a problem either.

Two things are not developmental milestones, despite the internet. Teething has its own timing, covered in the teething chart. Starting solids at around 6 months is a feeding decision with readiness signs of its own, covered in the 6-month solids guide.

Frequently asked questions

Is it normal for my baby to be late on one milestone?

Usually yes. The 75% rule means a quarter of children are later than the checklist age by design ¹, and the motor windows are wider still ³. What matters is the trend, and whether the other skills are moving.

At what age should a baby sit, crawl and walk?

In the WHO five-country sample, sitting without support runs from 3.8 to 9.2 months, hands-and-knees crawling from 5.2 to 13.5 months, and walking alone from 8.2 to 17.6 months ³. These are ranges, not appointments, and rolling is not one of the milestones the study measured.

Does early walking mean my baby is advanced?

It means their motor system matured early. It does not predict intelligence, and it is not something to train for ³. A baby who walks at 9 months and one who walks at 16 months are both ordinary.

Should I correct for prematurity?

Yes, until age 2. Developmental age is calculated from the due date rather than the birth date, which is the NHS rule and standard practice ⁶. Without that correction, a premature baby looks late on every checklist.

What happens if a screening test flags a concern?

The next step is a fuller evaluation, not a diagnosis. In the United States that usually means a referral to publicly funded early intervention services for children under 3; in the UK, a referral through the health visitor or GP. Early support is the reason the checkpoints exist at all ¹⁷.

Can a baby skip crawling?

Yes. In the WHO study, 4.3% of children never crawled on hands and knees ³. Bottom-shuffling and commando crawling are common alternatives, and the skill being built is movement, not a particular gait.

Disclaimer: This article is for informational purposes only and does not constitute medical advice. Milestone ages describe populations, not individual children, and they are not a substitute for developmental screening. If you are worried about any skill, raise it at the next visit or sooner; do not wait for a scheduled check to pass. Seek prompt review if a child loses a skill they already had.

CDC, “Key Points about CDC’s Developmental Milestone Checklists,” Learn the Signs. Act Early., updated 16 February 2026. Source ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³
Zubler JM, Wiggins LD, Macias MM, et al., “Evidence-Informed Milestones for Developmental Surveillance Tools,” Pediatrics, March 2022. Source ↩ ↩² ↩³
WHO Multicentre Growth Reference Study Group, “WHO Motor Development Study: windows of achievement for six gross motor development milestones,” Acta Paediatrica Supplement, April 2006. Source ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶
Zubler J, Whitaker T, “CDC’s Revised Developmental Milestone Checklists,” American Family Physician, October 2022. Source ↩ ↩² ↩³ ↩⁴ ↩⁵
CDC, “CDC’s Developmental Milestones,” Learn the Signs. Act Early., accessed October 2026. Source ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶
NHS, “Your baby’s health and development reviews,” reviewed 4 April 2023. Source ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷
Lipkin PH, Macias MM; Council on Children with Disabilities, Section on Developmental and Behavioral Pediatrics, “Promoting Optimal Development: Identifying Infants and Young Children With Developmental Disorders Through Developmental Surveillance and Screening,” Pediatrics, January 2020. Source ↩ ↩² ↩³ ↩⁴ ↩⁵
Hyman SL, Levy SE, Myers SM; Council on Children with Disabilities, Section on Developmental and Behavioral Pediatrics, “Identification, Evaluation, and Management of Children With Autism Spectrum Disorder,” Pediatrics, January 2020. Source ↩