Baby measuring big with gestational diabetes

The weight on a growth scan is an estimate, not a measurement, and scan weights are often off, more often too high than too low. NICE advises a growth scan every 4 weeks from 28 to 36 weeks with diabetes, so one scan is not the final word. Keeping your blood sugar in range is the part you control: in a trial, treating mild gestational diabetes cut the share of babies born over 4,000 grams by more than half.

Last reviewed 2026-10-21. Written by the First Fork team from the published guidelines listed below; not reviewed by a clinician. This page is not medical advice.

What "measuring big" means

A growth scan measures the baby's head, belly and thigh bone, and a formula turns these into an estimated weight.¹ Some words you may hear:

Word	What it means
Estimated fetal weight	The scan's guess at the baby's weight, from those measurements.¹
Above the 90th or 95th percentile	Bigger than 90 or 95 of 100 babies at the same week. Studies use this cut-off for the weight estimate or the belly measurement to call a baby big on a scan.²
Macrosomia	A birth weight over 4,000 grams (about 8 pounds 13 ounces).³
A "large" baby on ACOG's patient page	9 pounds (about 4 kg) or more.⁴

Why gestational diabetes can make a baby grow big

With gestational diabetes, your body passes more sugar to the baby than it needs, and the baby can gain a lot of weight.⁴ High blood sugar after meals is a main driver of this extra growth.⁵ Babies of mothers with diabetes are larger more often when the diabetes is not well controlled.⁶ A baby of 9 pounds or more can make labor harder and raises the chance of heavy bleeding and severe tears, and big babies are more likely to have a shoulder injury in a vaginal birth.⁴ For the other risks, see gestational diabetes risks for the baby.

How accurate is the weight on the scan?

It is an estimate. A Cochrane review says scan estimates of a baby's weight are often inaccurate, so many women may worry for no reason.³
It is more often too high. A 2018 review found that scan weights were most often higher than the real birth weight. In recent studies the usual error was under 10%, which is the margin counted as acceptable.¹ On an estimate of 4,000 grams, 10% is 400 grams (about 14 ounces).
Different people can get different numbers. A review found large differences when the same baby is measured by two people, or twice by the same person.⁷
With diabetes. A 2024 review of 20 studies found that scans picked up about 7 in 10 babies who were truly big, and flagged about 1 in 10 babies who were not. Scans within 2 weeks of birth were the most accurate.² In one US study, diabetes did not make scans near birth less accurate.⁸

How often growth is checked

NICE advises a scan of the baby's growth and the fluid around the baby every 4 weeks from 28 to 36 weeks for pregnant people with diabetes.⁹ Ultrasound in the last months of pregnancy can follow the baby's size.⁶ Your team may use a different schedule; ask when your next scan is.

What your team may do

Look at your numbers and medicine. NICE advises that the team consider insulin straight away for someone newly diagnosed with a fasting level of 6.0 to 6.9 mmol/L (about 108 to 124 mg/dL) and a large baby on the scan.⁹
Talk about how to give birth. If a scan shows a very large baby, NICE advises the team to explain the risks and benefits of a vaginal birth, an induction and a cesarean.⁹ If your ob-gyn thinks the baby is too big for a safe vaginal birth, you may discuss a scheduled cesarean.⁴ See C-section with gestational diabetes.
Talk about when to give birth. NICE advises considering birth before 40 weeks plus 6 days when there are problems for you or the baby.⁹ A Cochrane review of induction at or near term for a suspected big baby found fewer shoulder problems at birth (shoulder dystocia), fewer broken bones and a lower average birth weight, with no clear change in the cesarean rate. Because scan weights are often off, the review says many of these inductions may not be needed, and the choice should be talked through with parents.³ See induction and birth with gestational diabetes.

What you can do

Keep your readings in range. In a US trial of mild gestational diabetes, treatment cut births over 4,000 grams from 14.3% to 5.9%.¹⁰ NICE says good blood sugar control through pregnancy lowers the risk of a very large baby.⁹ Because high readings after meals drive extra growth,⁵ the checks after meals matter. See when to check after eating and blood sugar targets.

Bring your blood sugar record to every visit, as ACOG advises.⁴ In First Fork, the one-page clinic report shows each reading with its time and meal.

Questions to ask after a scan

What was the estimated weight, and what percentile is that?
Is the whole estimate big, or mostly the belly measurement?
How far off could this estimate be?
When is the next scan?
Does this change when or how you plan the birth?
Do the blood sugar targets or the medicine need to change?

Questions

How accurate is the baby's weight on a growth scan?

It is an estimate and is often off, more often too high than too low. In recent studies the usual error was under 10%. Scans within 2 weeks of birth were the most accurate.

What weight counts as a big baby?

A birth weight over 4,000 grams (about 8 pounds 13 ounces) is called macrosomia. ACOG's patient page calls a baby of 9 pounds (about 4 kg) or more large.

Does a big baby on the scan mean I need a C-section?

Not always. NICE advises your team to explain the risks and benefits of a vaginal birth, an induction and a cesarean. If your ob-gyn thinks the baby is too big for a safe vaginal birth, you may discuss a scheduled cesarean.

Can I stop my baby from getting too big?

Keeping your blood sugar in range lowers the chance. In a trial, treating mild gestational diabetes cut births over 4,000 grams from 14.3% to 5.9%.

How often will I have growth scans?

NICE advises a growth scan every 4 weeks from 28 to 36 weeks with diabetes. Your team may use a different schedule.

Sources

Milner J, Arezina J. The accuracy of ultrasound estimation of fetal weight in comparison to birth weight: a systematic review. Ultrasound 2018;26(1):32-41. https://pmc.ncbi.nlm.nih.gov/articles/PMC5810856/
Panunzi C, et al. Ultrasound prediction of fetal macrosomia in pregnancies complicated by diabetes mellitus: a systematic review and meta-analysis. J Perinat Med 2024;52(6):623-632. https://pubmed.ncbi.nlm.nih.gov/38860644/
Boulvain M, Thornton JG. Induction of labour at or near term for suspected fetal macrosomia. Cochrane Database Syst Rev 2023;3:CD000938 (systematic review). https://pmc.ncbi.nlm.nih.gov/articles/PMC9995561/
ACOG. Gestational Diabetes (FAQ 177, reviewed April 2026). https://www.acog.org/womens-health/faqs/gestational-diabetes
American Diabetes Association. 15. Management of Diabetes in Pregnancy: Standards of Care in Diabetes—2026. Diabetes Care 2026;49(Suppl. 1):S321. https://pmc.ncbi.nlm.nih.gov/articles/PMC12690181/
MedlinePlus (US National Library of Medicine). Infant of diabetic mother. https://medlineplus.gov/ency/article/001597.htm
Dudley NJ. A systematic review of the ultrasound estimation of fetal weight. Ultrasound Obstet Gynecol 2005;25(1):80-89. https://pubmed.ncbi.nlm.nih.gov/15505877/
Valent AM, et al. Accuracy of sonographically estimated fetal weight near delivery in pregnancies complicated with diabetes mellitus. J Ultrasound Med 2017;36(3):593-599. https://pubmed.ncbi.nlm.nih.gov/28108981/
NICE guideline NG3. Diabetes in pregnancy: management from preconception to the postnatal period (recommendations). https://www.nice.org.uk/guidance/ng3/chapter/Recommendations
Landon MB, et al. A multicenter, randomized trial of treatment for mild gestational diabetes. N Engl J Med 2009;361:1339–1348. https://pmc.ncbi.nlm.nih.gov/articles/PMC2804874/

See a mistake, or something unclear? Report a problem