Growth percentiles, and why the trend matters more
What a percentile actually says, why crossing lines is the signal rather than the level, and how WHO and IAP charts differ.
Last reviewed · 1,433 words
In short
- A percentile compares your child to a reference population. The 25th percentile means 25% of children are smaller — it is not a score.
- Following a percentile consistently is normal at any level. Crossing two major lines in either direction is what warrants attention.
- WHO charts to age five describe how children grow under optimal conditions. They are a standard, not an average.
- Breastfed and formula-fed infants grow differently after the first months, and using the wrong chart produces false alarms.
- Weight-for-height and BMI-for-age say more about nutrition than weight alone does.
A growth percentile places a child's measurement within a reference population of the same age and sex. At the 25th percentile, 25 out of 100 children that age are smaller and 75 are larger.
That is all it says. It is a position, not a grade, and almost every misreading of a growth chart comes from treating it as one.
The percentile is not a score
A child at the 15th percentile is not doing worse than one at the 85th. Both are within the normal range, and someone has to be at the 15th — by construction, 15% of perfectly healthy children are at or below it.
The charts are published as LMS parameters — a skew, a median and a coefficient of variation for each age — and the percentile comes from two steps:
z = ((measurement ÷ M)^L − 1) ÷ (L × S), or ln(measurement ÷ M) ÷ S when L is zero
percentile = Φ(z) × 100, Φ being the standard normal cumulative distribution.
WHO flags a z below −2 or above +2 as worth a clinical look, which is the 2nd and 98th percentile.
The range from the 3rd to the 97th percentile covers 94% of children. Genetics is the dominant input: short parents have short children, and a child at the 10th percentile with parents at the 10th is exactly where they should be.
What matters is not the number but its stability.
Crossing lines is the signal
Following a percentile is normal. A child tracking the 25th from six months to five years is growing well, whatever the level.
Crossing two major lines is worth investigating. The major lines are 3, 15, 50, 85 and 97. Falling from the 75th to the 25th, or climbing from the 15th to the 75th, is a change in the growth pattern rather than a change in size.
Two exceptions where crossing is expected:
The first two years. Birth size reflects the intrauterine environment; by about 18 months a child settles onto their genetic trajectory. Crossing lines during this period is common and usually normal.
Puberty. Growth spurts start at different ages, so a child who starts late will cross downward and then recover.
Outside those, a sustained change in trajectory is the finding a clinician acts on.
WHO against IAP charts
Two sets are used in India and they differ in what they represent.
WHO standards, birth to five years, were built from children raised under optimal conditions across six countries including India — breastfed, well nourished, non-smoking mothers, good healthcare. They describe how children should grow, and they are prescriptive rather than descriptive.
IAP charts, five to eighteen years, were built from Indian children and describe how Indian children do grow. They are used above five because the WHO standards end there and because adolescent growth patterns are more population-specific.
The distinction matters. A prescriptive standard identifies undernutrition that a descriptive chart, drawn from a population where undernutrition is common, would normalise. The current Indian practice — WHO to five, IAP thereafter — is a considered compromise between the two purposes.
Note that the IAP revised its charts in 2015 partly because the earlier ones, drawn from affluent urban children, were being used to normalise rising overweight.
Breastfed and formula-fed infants grow differently
Breastfed infants typically gain faster in the first two to three months and slower from four to twelve months than formula-fed infants.
Because the WHO standards are based on breastfed infants, a formula-fed baby may appear to climb percentiles on them. Conversely, using an older chart based on formula-fed populations makes a healthy breastfed baby look as though they are faltering at around six months — which has historically prompted a great deal of unnecessary supplementation.
Any apparent slowdown at four to six months in an otherwise thriving breastfed baby is worth checking against the right chart before acting on.
Which measurement to use
Weight-for-age is the most commonly tracked and the least informative on its own, because it cannot distinguish a small child from an undernourished one.
Length or height-for-age reflects long-term nutrition and health. Persistently low height-for-age is stunting, which indicates chronic undernutrition.
Weight-for-height reflects current nutritional state. Low weight-for-height is wasting, which indicates acute undernutrition and needs prompt attention.
BMI-for-age, used above two years, is the standard measure for identifying overweight in children. Adult BMI thresholds do not apply — children are assessed on age-and-sex percentiles.
Head circumference, to about three years, tracks brain growth. A crossing pattern here is taken more seriously than elsewhere.
The pair worth watching together is height-for-age and weight-for-height. A child low on both is in a different situation from one low on height alone, and only one measurement cannot tell them apart.
Measuring accurately
Bad measurements produce false alarms and false reassurance in equal number.
Length lying down until two years, standing height after. The two differ by about 0.7 cm, which is enough to look like a change if the method switches without noting it.
Weight without clothes or nappy for infants, in minimal clothing for older children, on the same scale each time.
Head circumference at the widest point, above the eyebrows and around the back of the head.
The same time of day, since a child weighs more in the evening and is measurably taller in the morning.
Measure monthly to six months, every two months to a year, and every three to six months thereafter. Measuring more often than that produces noise rather than information, and the noise causes worry.
When to raise it
Crossing two major percentile lines in either direction, sustained across visits.
Below the 3rd percentile on any measure, or a sudden change on any.
Weight loss, or no weight gain over two months in an infant.
Head circumference crossing lines, in either direction.
A large gap between the height and weight percentiles, particularly weight well below height.
Any concern alongside the numbers — poor feeding, low energy, developmental delay, frequent illness. The chart supports a clinical judgement; it does not replace one.
Growth faltering has many causes, most of them treatable: feeding difficulty, coeliac disease, cow's milk protein allergy, recurrent infection, thyroid problems, or simply insufficient intake. Identifying it early is the point of the chart.
Premature babies and corrected age
A baby born early is plotted against corrected age rather than the age since birth.
A baby born at 32 weeks is eight weeks premature. At four months old chronologically, they are plotted as a two-month-old — and they should be compared to two-month-olds, not four-month-olds.
The correction is applied until about two years for growth, and up to three for head circumference and developmental milestones. By then the difference has usually resolved and the child is plotted normally.
Failing to correct produces a false picture of growth faltering, and it is a common error in charts kept at home. Very premature babies are usually plotted on specialist preterm charts such as the Fenton chart until they reach term-equivalent age, and moved to the standard charts afterwards.
Catch-up and catch-down growth
Two normal patterns look alarming and are not.
Catch-down. A baby born large, often to a mother with gestational diabetes, frequently drifts downward across percentiles over the first year as they settle onto their genetic trajectory. Birth size reflects the womb; the trajectory afterwards reflects genetics.
Catch-up. A baby born small for gestational age often climbs, sometimes substantially, over the first two years. Around 85% of small-for-dates babies catch up by two, and those who have not by then are usually referred.
Both are expected to settle by about 18 months to two years, after which the child should follow a consistent percentile. Crossing lines after that point is the observation that warrants attention.
The reliable interpretation of any chart therefore depends on knowing where the child started and how long ago. A single plotted point without that history is close to uninterpretable, whatever percentile it lands on.
What this calculator assumes
- WHO growth standards to five years and IAP references above, which is standard Indian practice.
- Age in completed months, and corrected age for premature infants up to two years — a baby born eight weeks early is plotted eight weeks back.
- Accurate measurements taken as described above.
- Single measurements. Plotting a series over time is what the chart is for, and this page gives you one point on it.
- This is not a diagnosis. Growth concerns belong with a paediatrician who can see the child.