Fetal growth percentile calculator

Three standards, three different answers

A fetal growth centile is not a property of a baby. It is a property of a baby and a chart. The same estimated weight at the same gestational age can be the 39th centile on one published standard and the 71st on another, and this calculator shows all three at once rather than picking one and hoping you do not notice.

Hadlock 1991

Hadlock FP, Harrist RB, Martinez-Poyer J, “In utero analysis of fetal growth: a sonographic weight standard”, Radiology 1991;181(1):129–133 (PMID 1887021). The paper prints the model in running text:

“The optimal model was a natural log model of weight in grams on menstrual age (in weeks) and menstrual age squared: Log n weight (g) = 0.578 + 0.332 MA − 0.00354 MA², where MA is menstrual age (standard deviation = 0.12, R² = 99.1%).”

So the median weight at a gestational age of GA weeks is:

median weight (g) = exp(0.578 + 0.332 × GA − 0.00354 × GA²)

At 32 weeks that is exp(0.578 + 10.624 − 3.62496) = 1953 g, which is exactly the figure printed in the paper’s own Table 1. At 40 weeks it is 3619 g, again matching the paper.

For the centile, this calculator uses a constant coefficient of variation of 13.3%:

z = (EFW ÷ median − 1) ÷ 0.133

That is the figure that reproduces the paper’s printed centile table. Read Table 1 back and every row sits at exactly 0.83 and 1.17 times the median for the 10th and 90th centiles, and 0.75 and 1.25 for the 3rd and 97th — which, under the normal distribution, is a coefficient of variation of 13.3% (1.2816 × 0.133 = 0.1705; 1.8808 × 0.133 = 0.2502). The paper’s prose separately quotes ±12.7% for the raw residual scatter, which is slightly narrower; the calculator reports the centile both ways so the difference is visible. Range: the paper calculated values from 10 to 41 weeks.

INTERGROWTH-21st

Stirnemann J, Villar J, Salomon LJ, Ohuma E, Ruyan P, Altman DG, et al., “International estimated fetal weight standards of the INTERGROWTH-21st Project”, Ultrasound in Obstetrics & Gynecology 2017;49(4):478–486 (doi 10.1002/uog.17347). A Box-Cox skewed distribution fitted to the natural log of estimated fetal weight, published for 22 to 40 weeks. With GA in exact weeks:

• μ(GA) = 4.956737 + 0.0005019687 × GA³ − 0.0001227065 × GA³ × ln(GA)
• σ(GA) = 10⁻⁴ × (−6.997171 + 0.057559 × GA³ − 0.01493946 × GA³ × ln(GA))
• λ(GA) = −4.257629 − 2162.234 × GA⁻² + 0.0002301829 × GA³
• z = ((ln(EFW) ÷ μ)^λ − 1) ÷ (σ × λ)

Run backwards, those equations regenerate the published INTERGROWTH-21st centile table — the 3rd, 5th, 10th, 50th, 90th, 95th and 97th centiles for every week from 22 to 40 — to within 0.8% at every one of its 133 cells and within 0.2% at 128 of them. At 22 weeks the equations give a median of 525 g against a published 525 g; at 40 weeks, 3338 g against a published 3338 g.

The same group also publishes an estimated fetal weight equation of its own, from head and abdominal circumference only: ln(EFW) = 5.084820 − 54.06633 × (AC/100)³ − 95.80076 × (AC/100)³ × ln(AC/100) + 3.136370 × (HC/100), with AC and HC in centimetres.

WHO Fetal Growth Charts 2017

Kiserud T, Piaggio G, Carroli G, Widmer M, Carvalho J, Neerup Jensen L, et al., “The World Health Organization Fetal Growth Charts: A Multinational Longitudinal Study of Ultrasound Biometric Measurements and Estimated Fetal Weight”, PLoS Medicine 2017;14(1):e1002220.

WHO fitted its curves by quantile regression rather than to a distribution, so it published centiles rather than an equation. There is no formula to quote; the numbers themselves are the standard. Table 11 of that paper, “Growth chart for estimated fetal weight regardless of fetal sex”, gives the 2.5th, 5th, 10th, 25th, 50th, 75th, 90th, 95th and 97.5th centiles for every week from 14 to 40, and this calculator carries that table verbatim and places a weight between the published centiles on the z-scale. WHO calculated its own estimated weights by “including HC, AC, and FL in Hadlock et al.’s third formula”. Selected medians: 330 g at 20 weeks, 1189 g at 28 weeks, 1901 g at 32 weeks, 3617 g at 40 weeks. WHO also published sex-specific tables and found male fetuses larger by 3.5% at the lower quantiles and 4.5% at the upper ones; this tool uses the sex-combined table because fetal sex is often unknown at a growth scan.

Why the choice of chart is a live safety question

On 10 December 2025, NHS England and the Royal College of Obstetricians and Gynaecologists wrote to Trust medical directors, chief nurses, clinical directors for maternity services and heads of midwifery across England (publication reference PRN02254):

“Analysis indicates that Intergrowth EFW charts can lead to a lower screen positive rate of small for gestational age (SGA) fetuses (<10th centile). This means that some fetuses that otherwise would have undergone enhanced surveillance have not.”

Trusts using any INTERGROWTH estimated fetal weight chart were told to stop and move to one of “Hadlock formula”, “World Health Organization (WHO)”, “Fetal Medicine Foundation (FMF)” or “Perinatal Institute GROW”, with the change “completed by 31 March 2026”. During the crossover, the letter told units that “any EFW which is below the 25th percentile on an Intergrowth chart should be manually checked against the publicly available WHO thresholds”. The letter is explicit that this concerns estimated fetal weight charts only — symphysis-fundal height charts were not affected — and it also notes that “some EFW charts (for example, Hadlock and WHO) end at 40 weeks gestation”.

Customised charts, which this tool cannot reproduce

The fourth family of charts is customised: GROW (gestation related optimal weight, from the Perinatal Institute) and the Fetal Medicine Foundation standard adjust the expected weight for the mother’s height, weight at booking, parity and ethnic group, so two babies with identical scans get different centiles. Many UK maternity units use them and NHS England names them as acceptable alternatives. Their coefficients are not openly published, so this calculator tells you they exist and will give a different answer rather than guessing at them. Customised centiles are available from gestation.net.

The uncertainty that dwarfs all of this

Before the weight ever reaches a chart, it was estimated by ultrasound, with a random error of about 7.5% at one standard deviation (Hadlock et al. 1985) — roughly ±15% at 95% confidence. A 2000 g estimate is really somewhere between about 1700 g and 2300 g, and at 34 weeks that window alone spans the 2nd centile to the 40th on the Hadlock chart. This calculator shows that range in the result panel. It is usually the largest single source of uncertainty in the answer, larger than the disagreement between the charts.