Baby growth percentile calculator
Enter your baby’s date of birth, sex and measurements and this tool gives you the weight, length and head circumference centiles on the UK-WHO, WHO or CDC chart — and on the other two at the same time, so you can see how much the chart matters. It runs in your browser.
Baby growth percentile calculator
No sign-up · PrivateEvery growth chart is drawn separately for boys and girls, so this is not optional — the same weight sits on a different centile depending which of the two it is plotted on.
The day your baby was born.
Leave blank for today. Use the day of the weigh-in, not the day you are reading the result.
Leave blank if your baby was born at term. Below 37 weeks the tool corrects the age before plotting, because plotting a preterm baby against uncorrected age invents a low centile out of arithmetic.
In kilograms — 7.5, not 7500. Type grams and the tool will convert them and tell you it has.
In centimetres. Babies are measured lying down; from 2 years the charts expect a standing height, which is about 0.7 cm shorter.
In centimetres, around the widest part. Sometimes written OFC, for occipitofrontal circumference.
Choose boy or girl first. Growth charts are drawn separately for each, and a centile means nothing without knowing which chart it came from.
Nothing you type leaves your device. The whole calculation runs in your browser.
How this is calculated
FormulaHow a centile is worked out
All three charts here use the LMS method (Cole TJ, Green PJ, “Smoothing reference centile curves: the LMS method and penalized likelihood”, Statistics in Medicine 1992;11(10):1305–1319, PMID 1518992). For every age and sex the reference publishes three numbers: L, a skewness parameter; M, the median; and S, the coefficient of variation. A measurement x becomes a z-score:
- z = ((x ÷ M)L − 1) ÷ (L × S) when L is not zero
- z = ln(x ÷ M) ÷ S when L is zero
The centile is then the standard normal cumulative probability of that z-score. A worked example, using WHO’s own published weight-for-age parameters for boys at day 182 (exactly 26 weeks): L = 0.1260, M = 7.9260 kg, S = 0.10959. For a baby weighing 7.5 kg, (7.5 ÷ 7.9260) = 0.94625; raised to the power 0.1260 = 0.99306; minus 1 = −0.00694; divided by (0.1260 × 0.10959 = 0.013808) = z = −0.502, which is the 31st centile.
Which reference data, and where it comes from
WHO Child Growth Standards 2006 — birth to 5 years, boys and girls, for weight-for-age, length/height-for-age and head-circumference-for-age. The L, M and S values used here are WHO’s own published expanded z-score tables, downloaded from cdn.who.int. These are a standard rather than a reference: they describe how healthy, breastfed children of non-smoking, well-nourished mothers in Brazil, Ghana, India, Norway, Oman and the United States did grow, on the argument that this is how children in general should be able to grow.
UK-WHO (RCPCH) — the chart in the red book, and a deliberate hybrid. UK90 (British 1990) reference data runs from 23 weeks of gestation to 2 weeks of postnatal age; the WHO standards run from 2 weeks to 4 years; UK90 resumes from 4 years. RCPCH documents the middle section as “the WHO standard for weight, BMI and head circumference from 2 weeks to 4 years, for length 2 weeks to 2 years and height 2–4 years… It is exactly the same data as the LMS data included in the Z-score tables accessed from the WHO website, except there is no birthweight.” That is the section this calculator implements. Below 2 weeks of age it declines to plot, for two reasons: it does not carry UK90 birth coefficients it could verify from a primary source, and a birth measurement belongs on the birth centile section of the chart, plotted against gestational age at birth rather than against postnatal age — a different calculation entirely.
CDC 2000 — the US national reference, from cdc.gov/growthcharts. The L, M and S values are taken from CDC’s published data files: wtageinf.csv, lenageinf.csv and hcageinf.csv for birth to 36 months, and wtage.csv and statage.csv for 2 to 20 years. Unlike WHO, this is descriptive: it records how US children actually grew, largely from national survey data, in a period when formula feeding was much more common. Re-deriving CDC’s own published 3rd, 10th, 50th, 90th and 97th centile columns from those L, M and S values agrees to within 0.23%, which is the check that the implementation here is faithful.
CDC’s own recommendation is to use the WHO standards for children under 2 years and the CDC charts from 2 to 20 years. In the UK the answer is simpler: use UK-WHO.
Interpolation, and the step at 24 months
WHO publishes L, M and S for every single day from 0 to 1856. Carrying 1857 rows for each of six tables in a web page would be absurd, so this calculator stores 100 anchor days — every day to 21 days, every week to 26 weeks, every month to 60 months, plus day 730 — and interpolates linearly between them. Checked against all 1827 published daily rows at z-scores from −3 to +3, the worst centile this approximation gets wrong anywhere in any of the six tables is out by 0.47 of a centile point. CDC publishes its parameters at half-month ages and those are used directly, interpolated linearly, which is what CDC’s own program does.
Day 730 is an anchor on purpose. WHO drops the median length/height by 0.7 cm at 24 months, where measurement switches from lying down to standing up. That step is real and must not be smoothed away by interpolation.
Premature babies
Plotting a preterm baby against uncorrected age manufactures a low centile out of arithmetic and frightens people for no reason. If a gestational age at birth below 37 weeks is entered, this calculator subtracts the weeks of prematurity from the age before plotting, and then shows what the uncorrected age would have produced so the difference is visible. A boy born at 30 weeks and weighing 5.5 kg at 179 days old is on the 3.9th centile at his corrected age of 109 days; plotted at his uncorrected age he appears below the 0.1st. Nothing about him changed.
The correction follows the RCPCH convention: correct until 1 year for babies born at 32 to 36 weeks, and until 2 years for babies born before 32 weeks. ShePrep’s adjusted age calculator works through the arithmetic in full.
Why the centile lines look different in different countries
UK charts are printed with nine lines — the 0.4th, 2nd, 9th, 25th, 50th, 75th, 91st, 98th and 99.6th — each two thirds of a standard deviation apart. US charts are printed at the 3rd or 5th, 10th, 25th, 50th, 75th, 90th, and 95th or 97th. So a British parent is told “on the 9th centile” and an American parent “on the 10th” about very nearly the same line, and neither child is different from the other. This calculator prints the lines belonging to whichever chart you have selected.
What a centile is not
It is not a score, a grade or a target, and there is no centile a baby is supposed to be on. Half of all babies are below the 50th, by definition. Two in a hundred are below the 2nd, by definition. What health visitors and paediatricians watch is the shape of the line over months — whether a baby is tracking along their own centile — not the height of any single point on it. And measurement error is larger than people expect, particularly for length: a centimetre at six months is worth several centile points, which is exactly why nobody acts on one reading.
The weigh-in
You go to the clinic, your baby is put on the scales, someone draws a dot in the red book, and then says a number. Ninth centile. Twenty-fifth. Second. And whatever they say next, the number is what you take home and turn over at three in the morning.
So here is the number, said properly. A centile is a position in a queue of a hundred babies the same age and sex. The 9th centile means eight would weigh less and ninety-one would weigh more. It is not a mark out of a hundred. There is no pass. Two babies in every hundred are below the 2nd centile and they are supposed to be, because that is what the 2nd centile means.
The line matters, the dot does not
What your health visitor is actually looking at is not the dot. It is the shape of the line the dots make. A baby who has tracked steadily along the 9th centile since birth is a small baby growing normally. A baby who has come down from the 75th to the 9th over two months is a different conversation, even though today’s dot is in the same place.
This is why a single measurement almost never means anything on its own, and why the answer to “is this bad?” is nearly always “what did the last three look like?”
Three charts, three different answers
UK-WHO, WHO and CDC are not three renderings of the same data. They are different reference populations, collected differently, for different purposes.
The WHO standards describe how healthy, breastfed babies of well-nourished mothers grew, in six countries, and they were built deliberately as a picture of how children should be able to grow. The CDC charts describe how American children actually did grow, mostly from national survey data collected when formula feeding was far more common. The UK-WHO chart is a hybrid: UK birth data for the first fortnight, then the WHO standards to four years.
Those differences show. A 3.1 kg newborn girl at three days old is on about the 39th centile on the WHO chart and the 23rd on CDC. Formula-fed and breastfed babies gain weight on different curves through the first year, and the two references were built on different mixtures of the two, so the gap is widest exactly where parents worry most.
The right response to that is not to average them and it is certainly not to hunt for the chart that gives the nicest number. It is to use the one your health service uses, so that today’s dot is comparable with the last one. In the UK that is UK-WHO. In the US it is WHO under two and CDC from two.
If your baby was born early
Correct the age. This is the single most common way a growth chart gets read wrong, and it is a cruel one, because the arithmetic invents a problem that does not exist.
A baby born at 30 weeks was ten weeks early. At six months old by the calendar, they are three and a half months old by their due date, and they should be compared with three-and-a-half-month-olds. Plot them against six months and they will look alarmingly small — not because anything is wrong, but because you are asking a fourteen-week-old to weigh what a twenty-six-week-old weighs.
This calculator will do the correction for you if you enter the gestational age at birth, and it shows you both numbers so you can see the size of the error. ShePrep’s adjusted age calculator for premature babies explains the whole thing, including how long correction is usually applied.
Measurement error, especially length
Babies do not lie straight. They kick, they curl, they object. Length is the least reliable measurement on the chart, and a centimetre at six months is worth several centile points. Head circumference has the same problem in miniature — the tape has to sit in exactly the right place.
So when a single reading looks out of step with the ones either side of it, the most likely explanation by a wide margin is the measuring, not the growing. Good practice is to re-measure before reacting, and that is what a health visitor who shrugs at one odd dot is doing.
When to actually ask someone
Crossing downwards through two centile lines, weight that has stopped increasing, a baby who is feeding poorly or seems unwell, or a head circumference that is changing direction — those are worth a call, not because they are necessarily serious but because they are the things a person needs to look at rather than a page.
Take the chart with you. Take the corrected age if your baby was early. And if you are worried at all, that is reason enough — your health visitor, GP or paediatrician can see the whole line and the whole child, and nobody has ever minded being asked.
Sources
- WHO Child Growth Standards — z-score expanded tables (weight-for-age, length/height-for-age, head circumference-for-age), birth to 5 years — World Health Organization, accessed
- WHO Child Growth Standards — weight-for-age, downloadable L/M/S tables — World Health Organization, accessed
- UK-WHO growth charts, 0-4 years — Royal College of Paediatrics and Child Health, accessed
- RCPCH Digital Growth Charts: reference data (which datasets the UK-WHO chart uses, and at which ages) — Royal College of Paediatrics and Child Health, accessed
- CDC growth charts: percentile data files with LMS values (wtageinf.csv, lenageinf.csv, hcageinf.csv, wtage.csv, statage.csv) — Centers for Disease Control and Prevention (National Center for Health Statistics), accessed
- Smoothing reference centile curves: the LMS method and penalized likelihood. Stat Med 1992;11(10):1305–1319 — Cole TJ, Green PJ (PubMed, PMID 1518992), accessed