An estimated fetal weight (EFW) percentile tells you where your baby’s predicted size falls compared to other babies at the same gestational age. If your ultrasound report says “35th percentile,” it means the estimated weight is larger than roughly 35 percent of fetuses at that stage and smaller than the remaining 65 percent. The number itself is not a grade or a health score, though it becomes clinically meaningful when it falls near the extremes or changes sharply between scans.
How the Estimate Is Calculated
During a growth ultrasound, the sonographer measures several parts of the baby’s body. The key measurements typically include the circumference of the head, the circumference of the abdomen, and the length of the thighbone. These measurements get plugged into a mathematical formula that converts them into an estimated weight in grams. Multiple formulas exist, and the one your clinic uses can actually shift the result. A Swedish study evaluating three commonly used formulas found that each produced somewhat different weight estimates from the same set of measurements, which means the percentile you see on your report is partly a product of which formula your hospital selected.
That estimated weight is then plotted on a growth chart, which converts the raw gram figure into a percentile for the baby’s gestational age. The chart chosen for this step matters just as much as the formula, a point that causes real confusion for patients who switch providers or compare results across different health systems.
How Accurate Is the Number on Your Report
Ultrasound weight estimates are useful approximations, not precision instruments. A large study looking at how well these estimates matched actual birth weights found that the average error was about 2.4 percent, which sounds reassuring. But the individual spread around that average was wide. More revealingly, the accuracy depended heavily on where in the weight spectrum the baby fell. For babies in a mid-range weight category, the error was smallest. For babies at the low end, the method tended to overestimate weight by around 5 percent, while for the heaviest babies it underestimated weight by roughly 12 percent.1PubMed Central. Percent error of ultrasound examination to estimate fetal weight at term in different categories of birth weight with focus on maternal diabetes and obesity That pattern is worth knowing: ultrasound tends to be least reliable at the very extremes where the percentile carries the most clinical weight.
Another study examining third-trimester accuracy reported that about 86 percent of estimates fell within 15 percent of the actual birth weight at 37 to 40 weeks, and about 68 percent fell within 10 percent.2Obstetrics & Gynecology. Accuracy of Estimated Fetal Weight in Third Trimester So if your baby’s estimated weight is 2,500 grams, the actual weight at birth could reasonably be anywhere from about 2,125 to 2,875 grams. That’s a range wide enough to shift a percentile ranking by ten or twenty points in either direction. The operator performing the scan matters too: research on inter-observer variability found that different sonographers measuring the same fetus can produce results that differ by 1 to 3 percent, and even the same sonographer repeating a scan may vary by 1 to 2 percent.3PubMed Central. Estimating fetal weight for best clinical outcome
A separate issue is that the formulas themselves assume certain proportions between the baby’s head, belly, and legs. When a baby’s proportions differ from the norm, as happens in some growth-restricted fetuses whose heads are relatively spared while their abdomens are thin, the formula can produce misleading results. One analysis showed that fetal proportions meaningfully affected the error of the weight estimate and could alter the apparent growth trajectory depending on which formula was used.4PubMed Central. Caveats in the monitoring of fetal growth using ultrasound estimated fetal weight
Why the Choice of Growth Chart Changes Your Percentile
Once your baby’s estimated weight is calculated, it has to be compared against something. That “something” is a growth chart, and not all growth charts define the 10th or 90th percentile in the same place. There are two broad approaches. Population-based charts are built from large datasets of births in a particular country or region, treating all babies the same regardless of the mother’s characteristics. Customized charts adjust the expected range for factors like the mother’s height, weight, ethnic background, and whether she has had babies before.
The debate between these approaches is not academic hairsplitting. A systematic review comparing the two found that a single population-based chart may not adequately capture risks for adverse outcomes across different groups of mothers and babies.5Malaysian Journal of Public Health Medicine. Customised Fetal Growth Charts Versus Population-Based Growth Charts in Identifying Adverse Fetal Outcomes: A Systematic Review A separate study testing the assumptions behind customized charts against national birth data also found discrepancies between routine clinical EFW data and the customized model’s predictions.6PubMed Central. Testing the assumptions of customized intrauterine growth charts using national birth studies
A striking illustration comes from a comparison of customized charts against the INTERGROWTH-21st standard, which was designed as a universal chart based on healthy pregnancies across eight countries. When applied to ten different national birth cohorts, the INTERGROWTH-21st standard produced wildly different rates of small-for-gestational-age (SGA) classification, ranging from about 3 percent to nearly 17 percent depending on the population. That variation tracked closely with known physiological differences in birth weight across populations.7PubMed. Customized vs INTERGROWTH-21st standards for the assessment of birthweight and stillbirth risk at term In plain terms, a baby labeled “small” on one chart might be perfectly average on another, and the difference can come down to which country’s data the chart was built from.
What Drives Fetal Size in the First Place
Before worrying about what a percentile means, it helps to understand that healthy babies come in a wide range of sizes. Several non-pathological factors push a baby higher or lower on the growth curve. Maternal height is one of the strongest predictors. A study analyzing birth weight data found that for every additional centimeter of maternal height, newborn weight increased by about 7 grams.8PubMed. Maternal height, and ethnicity and birth weight: A retrospective cohort study of uncomplicated term vaginal deliveries in Malaysia That may not sound like much per centimeter, but across the full range of adult female height, the cumulative difference is substantial.
A related analysis found that on standard percentile charts, babies born to taller mothers had low rates of SGA and high rates of being classified as large for gestational age (LGA), while the reverse was true for shorter mothers. When the data were stratified by maternal height, mothers of similar stature from different ethnic groups showed similar average birth weights, SGA rates, and LGA rates.9PubMed. Maternal body height is a stronger predictor of birth weight than ethnicity: analysis of birth weight percentile charts In other words, much of what looks like ethnic variation in birth weight may actually reflect differences in average maternal height across populations. Other factors like maternal age, pre-pregnancy weight, parity, the baby’s sex, and gestational age at delivery all play a role too.
When a Low Percentile Signals a Problem
A baby whose estimated weight falls below the 10th percentile is generally classified as small for gestational age. But being SGA is a description, not a diagnosis. Some of these babies are constitutionally small, meaning they are healthy and simply take after smaller parents. The clinical concern arises when the small size is caused by fetal growth restriction (FGR), which means the baby is not reaching its genetic growth potential because something is going wrong.
The most common cause of FGR is placental insufficiency, a condition in which the placenta does not deliver enough blood, oxygen, and nutrients to the fetus. The underlying process involves poor development of the blood vessels connecting the placenta to the uterine wall. This leads to increased resistance to blood flow through the placenta and reduced nourishment reaching the baby.10PubMed Central. Placental insufficiency and fetal growth restriction When the placenta underperforms, the fetus responds by redirecting blood flow to protect the brain at the expense of other organs, particularly the abdomen. This redistribution can be detected on Doppler ultrasound.
A prospective study tracked fetuses that showed slowing growth trajectories and found a significant link between blood flow redistribution toward the brain and a decline in growth percentile between 28 weeks and birth.11PLoS ONE. Placental Insufficiency in Fetuses That Slow in Growth but Are Born Appropriate for Gestational Age: A Prospective Longitudinal Study Some of these babies ended up being born at a technically “normal” weight but had still experienced genuine placental insufficiency along the way. That finding underscores an important point: a single percentile measurement can miss problems that only become visible when growth is tracked over time.
The Trajectory Matters More Than One Number
Clinicians pay close attention to whether a baby’s percentile is stable, rising, or falling across serial scans. A fetus that consistently tracks along the 15th percentile from 20 weeks onward is in a very different clinical situation from one that starts at the 50th percentile and drops to the 15th by the third trimester. The first baby may simply be constitutionally small. The second baby is crossing percentile lines downward, which can signal that something has changed in the intrauterine environment.
Growth charts serve as the visual tool for this tracking. Clinicians use them to interpret the growth trajectory and calculate where the baby’s size falls at a given gestational age, then adjust monitoring plans accordingly, including the frequency of follow-up scans and the timing of delivery. The choice of chart can have a significant impact on how many fetuses get classified as growth restricted or large at certain gestational ages, which is why the ongoing debate about the optimal chart has real consequences for patient care.
Doppler Studies and Further Investigation
When the estimated weight percentile is low and growth restriction is suspected, providers typically add Doppler ultrasound studies to evaluate blood flow in the umbilical cord and the baby’s brain. Umbilical artery Doppler measures the resistance in the blood vessels connecting the baby to the placenta. Progressive worsening of this measurement, from mildly elevated resistance to absent or reversed blood flow during the heart’s resting phase, correlates with increasing placental failure and rising risk to the baby.12Lynchburg Journal of Medical Science. Umbilical Artery Doppler in Fetal Growth Restriction These Doppler patterns help clinicians decide how intensively to monitor the pregnancy and when to consider delivery.13PubMed. Umbilical artery Doppler in the assessment of fetal growth restriction
Middle cerebral artery Doppler adds another layer. When the baby diverts extra blood flow to the brain, this vessel shows decreased resistance, a finding sometimes called “brain sparing.” The ratio between the middle cerebral artery and umbilical artery Doppler readings helps distinguish between growth-restricted babies at higher and lower risk of complications.14Eurasian Journal of Critical Care. Role of Fetal Umbilical and Middle Cerebral Artery Doppler Indices in Determining Intrauterine Growth Restriction in Preeclamptic Pregnancies If your provider orders these additional tests after a low percentile reading, it does not automatically mean something is seriously wrong. It means they are gathering more information before deciding on a course of action.
When the Percentile Is High
A baby estimated above the 90th percentile is classified as large for gestational age, and one whose estimated weight exceeds about 4,000 grams (roughly 8 pounds 13 ounces) regardless of percentile may be described as macrosomic. The most well-known driver of excessive fetal growth is maternal insulin resistance. Higher blood sugar in the mother sends more glucose across the placenta to the baby, causing the fetus to produce extra insulin and deposit more fat and tissue. This mechanism explains the strong link between gestational diabetes, maternal obesity, and large babies.15Obstetrics, Gynaecology & Reproductive Medicine. Fetal macrosomia and large for gestational age
High percentiles carry their own set of concerns during labor and delivery, including shoulder dystocia, birth injuries, and a higher likelihood of cesarean section. But the accuracy problem cuts both ways here: recall that ultrasound tends to underestimate the weight of the heaviest babies. A baby truly near the top of the scale may be even larger than the report suggests, while a baby incorrectly estimated to be very large can lead to unnecessary interventions.
Long-Term Health Links at Both Extremes
Research over the past several decades has connected both unusually low and unusually high birth weight to health risks that extend well beyond the newborn period. Babies affected by fetal growth restriction face an increased risk of cardiovascular disease, metabolic syndrome, diabetes, respiratory problems, and chronic kidney disease later in life.16PubMed Central. Long-term implications of fetal growth restriction The prevailing explanation is that the fetus adapts to a nutrient-poor environment by programming its metabolism to be extremely efficient at storing energy, a strategy that becomes harmful when the child grows up with access to plentiful food.17PubMed Central. Fetal Growth Restriction and Its Metabolism-Related Long-Term Outcomes-Underlying Mechanisms and Clinical Implications
On the cardiovascular side specifically, growth restriction is associated with changes in the developing heart and blood vessels. These include altered heart shape, subtle dysfunction in how the heart muscle contracts, remodeling of artery walls, and impaired function of the inner lining of blood vessels. These changes, first detectable during fetal life, have been shown to persist into childhood and adolescence.18American Journal of Obstetrics and Gynecology. Long-term cardiovascular consequences of fetal growth restriction
At the other extreme, high birth weight and LGA status carry low to moderately elevated risks for certain childhood cancers, some psychiatric conditions, childhood hypertension, and both type 1 and type 2 diabetes. A systematic review and meta-analysis found that babies born weighing more than 4,000 grams had a moderately increased risk of Wilms tumor and a smaller increase in leukemia risk compared to babies under that threshold.19PubMed Central. The Association Between High Birth Weight and Long-Term Outcomes—Implications for Assisted Reproductive Technologies: A Systematic Review and Meta-Analysis These are population-level statistical associations, not individual predictions. A baby born at the 95th percentile is not destined for any particular health outcome, but the data reinforce the idea that extremes of fetal size can reflect underlying biological conditions that have lasting effects.
MRI as an Alternative for Estimating Weight
Ultrasound remains the standard tool for estimating fetal weight, but MRI is increasingly studied as an alternative for difficult cases. A comparison of the two methods found a striking difference in precision: over a quarter of ultrasound-based weight estimates had a relative error exceeding 10 percent, compared to only about 1 percent of MRI-based estimates. The range of agreement between estimated and actual weight was considerably narrower with MRI.20PubMed. Fetal weight estimation: comparison of two-dimensional US and MR imaging assessments MRI is not about to replace ultrasound for routine growth checks: it is more expensive, less widely available, and takes longer. But for pregnancies where the stakes of an inaccurate weight estimate are high, such as when delivery decisions hinge on whether the baby has crossed a critical size threshold, MRI offers a notably more precise measurement. Researchers are also exploring AI-based analysis of fetal MRI as a tool for earlier and more accurate detection of growth restriction and placental insufficiency.21Clinical and Experimental Obstetrics & Gynecology. Performance of a Deep Learning-Based Fetal Magnetic Resonance Imaging System in the Assessment of Fetal Growth Restriction and Placental Insufficiency