Your baby's size is estimated from three ultrasound measurements — head circumference, abdominal circumference, and femur length — combined into an estimated fetal weight with a typical margin of error of plus or minus 10 to 15 percent. At 20 weeks, a fetus averages roughly 10 inches (25 cm) crown to heel and about 10.6 ounces (300 g). By 40 weeks, the average is about 20 inches (51 cm) and 7.5 pounds (3.4 kg). Individual variation is wide and usually normal.
Those numbers answer the question at two bookends, but they hide the more useful story: how fetal size is measured, why the estimates are fuzzy, what growth curves actually compare, and when a size reading deserves attention rather than reassurance. This explainer goes a level deeper than the standard week-by-week chart.
The measurement mechanism behind every size estimate
Ultrasound does not weigh a fetus. It measures distances and infers weight through formulas. The sonographer Freezes an image, places calipers on anatomical landmarks, and software plugs the numbers into a regression equation derived from thousands of prior births. The most common is the Hadlock formula, which combines biparietal diameter (side-to-side skull width), head circumference, abdominal circumference, and femur length.
Each measurement carries its own error. Femur length reflects skeletal growth but varies with genetics and ethnicity. Abdominal circumference tracks liver size and subcutaneous fat, so it is the most sensitive marker of fetal nutrition — and the noisiest to capture, because a fetus who arches or breathes changes the plane mid-measurement. Head circumference is more reproducible but tells you less about overall nutritional status.
The composite estimated fetal weight (EFW) is therefore a statistical guess, not a scale reading. A 2021 analysis in a major obstetrics journal found that EFW falls within 10 percent of birth weight in roughly 70 to 80 percent of cases, and within 15 percent in about 90 percent. That means a reported 6-pound estimate at 36 weeks could reasonably be 5 pounds 4 ounces or 6 pounds 12 ounces. The formula is precise to the gram on paper and imprecise in the body. The plain version: the number on the screen is a well-educated estimate, not a fact.
Quantified context: what average size looks like across gestation
Fetal size follows a predictable trajectory, but the curve is not linear. Growth accelerates through the second trimester, peaks in the third, then plateaus in the final weeks as the placenta ages and room runs out. The table below shows typical averages; individual fetuses routinely fall above or below without any problem.
| Gestational age | Crown-heel length (avg) | Estimated weight (avg) | What is developing |
|---|---|---|---|
| 8 weeks | 0.6 in (16 mm) | ~0.04 oz (1 g) | Limb buds, early organ primordia |
| 12 weeks | 2.1 in (54 mm) | ~0.5 oz (14 g) | Fingers, kidneys producing urine |
| 20 weeks | 10 in (25 cm) | ~10.6 oz (300 g) | Movement felt, hearing begins |
| 28 weeks | 14.8 in (37.6 cm) | ~2.2 lb (1,000 g) | Eyes open, fat deposition starts |
| 32 weeks | 16.7 in (42.4 cm) | ~3.75 lb (1,700 g) | Rapid weight gain phase |
| 36 weeks | 18.7 in (47.4 cm) | ~5.8 lb (2,600 g) | Lungs maturing |
| 40 weeks | 20 in (51 cm) | ~7.5 lb (3,400 g) | Term; growth slows |
Two features matter more than any single row. First, weight gain in the third trimester is steep — roughly 200 to 250 grams per week around 32 to 36 weeks — so a two-week gap between scans can shift the estimate substantially. Second, length measurements become unreliable late in pregnancy because fetuses curl, so crown-rump gives way to crown-heel and the numbers get fuzzier. Growth curves plot these averages as percentiles, and a fetus at the 10th or 90th percentile is still within the normal band by definition.
Edge cases: when size readings diverge from the curve
Most size variation is constitutional. Small parents tend to have smaller babies; larger parents, larger ones. Genetics set a personal growth potential, and the ultrasound curve is a population average, not a prescription. A fetus tracking steadily along the 15th percentile is usually growing normally for that family.
Divergence becomes clinically interesting in three patterns. The first is a fetus measuring below the 10th percentile, which flags possible fetal growth restriction (FGR) — a real condition where the fetus does not reach its genetic potential, often from placental insufficiency. The second is a fetus above the 90th percentile, which raises the question of macrosomia, more common in pregnancies affected by diabetes. The third, and most telling, is a falling percentile across serial scans: a fetus who drops from the 50th to the 20th over several weeks is growing slower than expected even if both numbers look acceptable in isolation.
A counterintuitive detail professionals rely on: a single small measurement is often a dating error, not a growth problem. If a pregnancy is dated by last menstrual period rather than an early ultrasound, the gestational age could be off by a week or more, which shifts every percentile. Late ultrasound dating is unreliable for exactly this reason. The fix is an early scan, not a bigger baby.
Misconceptions about fetal size
The most persistent misconception is that bigger means healthier. It does not. Macrosomia carries its own risks — shoulder dystocia during delivery, neonatal hypoglycemia — and an above-average estimate is a reason for monitoring, not celebration.
A second misconception confuses fundal height with fetal weight. The tape measure a clinician uses from pubic bone to uterine top, in centimeters, roughly tracks gestational weeks between 20 and 36 weeks, but it reflects uterine size, amniotic fluid, and fetal position as much as fetal mass. A fundal height that measures two weeks behind is a screening flag for an ultrasound, not a diagnosis of a small baby.
A third: parents often compare estimates across pregnancies or with friends. EFW depends on the formula used, the sonographer's technique, and the machine's calibration. Comparing a 32-week estimate from one clinic to another is comparing two different measurement systems. The trend within one pregnancy, using consistent technique, is far more informative than any single number.
Practical actions: what to do with a size reading
Fetal size readings call for calibration, not panic. These steps translate an estimate into a decision.
Confirm gestational age with the earliest available ultrasound before interpreting any percentile; dating error is the most common cause of an "abnormal" size. Ask for the percentile and the trend, not just the weight — a stable 30th percentile is different from a falling one. Track serial scans if a reading is flagged; two measurements two to four weeks apart reveal growth velocity, which matters more than a single point. Discuss maternal factors (blood pressure, glucose tolerance, smoking) that affect placental function, since these drive most genuine growth problems. Treat the estimate as a range: a reported weight of 5 pounds 8 ounces plausibly means 5 to 6 pounds, and delivery planning should account for that spread.
Red-flag findings that warrant prompt clinical review:
A fetus measuring below the 3rd percentile, or below the 10th with abnormal umbilical artery Doppler. A percentile that falls across two or more consecutive scans. Fundal height measuring three or more centimeters behind gestational age. New maternal high blood pressure combined with a small-for-dates reading. A sudden drop in fetal movement alongside any size concern.
Reconciling the thresholds and the limits of what size can tell you
Three thresholds coexist in clinical practice and are easy to conflate. The 10th percentile defines small-for-gestational-age as a screening category. The 3rd percentile flags the highest-risk subgroup. And a drop of more than one quartile across serial scans defines abnormal growth velocity even when the absolute percentile stays above 10. These are not interchangeable: a fetus at the 8th percentile growing steadily may need nothing, while a fetus at the 25th percentile falling fast may need Doppler studies and closer surveillance. The number matters less than the trajectory and the context.
The deeper limitation is that ultrasound estimates fetal size, not fetal health. A constitutionally small fetus and a growth-restricted fetus can share a percentile, and only Doppler, amniotic fluid assessment, and maternal factors separate them. Similarly, EFW cannot reliably predict birth weight within a pound, so decisions about delivery mode based on size alone carry real uncertainty — a fact major obstetric bodies acknowledge when they caution against acting on estimated weight without corroborating findings.
What the science supports is this: fetal size is a statistical estimate with a known error band, best interpreted as a trend within one pregnancy, measured against an accurate gestational age, and contextualized by family genetics and maternal health. The question "how big is my baby" has an answer — several answers, in fact — but the most useful one is not a single number. It is a curve, a range, and a pattern, read together.
If you take one number away, take the gestational-age range rather than the point estimate, and ask your clinician where your baby sits on the curve and which way the curve is heading. That pairing does more work than any weight on a screen.