Gestational age (GA) is the number of weeks and days that have passed since the first day of your last menstrual period, while AUA stands for “average ultrasound age,” the estimate of how far along your pregnancy is based on ultrasound measurements of the baby. You will see both on nearly every prenatal report, and they do not always match. The gap between them tells your provider something important about whether the baby is growing as expected, whether your due date needs adjusting, or whether further monitoring is warranted.
How Gestational Age Is Calculated
GA starts counting from the first day of your last menstrual period (LMP), even though conception typically happens about two weeks later, around ovulation. This convention can feel strange, but it has been the standard for centuries, and the entire system of prenatal milestones, due dates, and delivery planning is built around it. The classic method for turning your LMP into an estimated due date is Naegele’s rule: take the first day of your last period, subtract three months, and add seven days. That gives you a due date roughly 280 days (40 weeks) from your LMP.
The catch is that this method assumes you have a textbook 28-day cycle with ovulation on day 14. Many people do not. Cycles that are longer, shorter, or irregular shift the actual day of ovulation and conception, which means the LMP-based GA can be off by days or even weeks. That is one reason ultrasound dating exists: it provides an independent check.
How AUA Is Determined
AUA comes from measuring the baby during an ultrasound and comparing those measurements to reference charts built from thousands of pregnancies. In the first trimester, the key measurement is crown-rump length (CRL), which is the distance from the top of the baby’s head to the bottom of its torso. CRL is remarkably consistent across healthy pregnancies in those early weeks, making it the most reliable single measurement for dating.
Once you move past the first trimester, the baby’s limbs, head, and abdomen grow at rates that vary more from one fetus to another. At that point, providers rely on a combination of measurements rather than a single one:
- Biparietal diameter (BPD): the width of the skull from one side to the other.
- Head circumference (HC): the distance around the baby’s head.
- Abdominal circumference (AC): the distance around the baby’s belly.
- Femur length (FL): the length of the thighbone.
Software in the ultrasound machine takes these measurements and calculates an average gestational age from them, which is what shows up as “AUA” on your report. Some formulas simply average the individual age estimates from each measurement, while others use regression equations that weight the measurements differently. A study comparing these approaches found that both methods perform similarly, though newer regression-derived formulas using national fetal growth data aim to improve precision in diverse populations.1PubMed. Estimating Gestational Age With Sonography: Regression-Derived Formula Versus the Fetal Biometric Average
Research has also explored whether other measurements can improve accuracy in the third trimester, when the standard four parameters become less reliable for dating. One study found that the transcerebellar diameter, a measurement of a structure at the back of the brain, correlated with LMP-based GA more strongly than BPD, AC, or FL, and remained useful even in cases of abnormal growth.2International Journal of Reproduction, Contraception, Obstetrics and Gynecology. Comparison of transcerebellar diameter with conventional fetal biometric parameters like biparietal diameter, head circumference, femur length and abdominal circumference in predicting gestational age in third trimester
Why GA and AUA Do Not Always Match
If your LMP-based GA says you are 20 weeks but the ultrasound says the baby looks like 21 weeks, or 18 weeks, that discrepancy has to mean something. The question is what. There are a few common explanations.
The simplest is that the LMP date is wrong. You may not remember the exact start of your last period, your cycle may be irregular, or you may have had spotting early in pregnancy that was mistaken for a period. All of these shift the LMP-based GA. A study comparing LMP-based dating with first-trimester ultrasound in women who reported certain periods found meaningful discrepancies even in that group, underscoring how often memory and cycle irregularity introduce error.3Journal of the Medical Association of Thailand. Discrepancy Rate and Associated Factors of Gestational Age Determination by Last Menstrual Period Compared with First-Trimester Ultrasound in Certain Menstrual Period Pregnant Women
A subtler cause is variation in the timing of implantation. Even when ovulation and conception happen on a known day, the embryo does not always implant at the same rate. Research has shown that early implantation leads to a larger CRL at the first-trimester scan, while late implantation leads to a smaller CRL, regardless of how fast the embryo is actually growing afterward. Implantation timing is a major driver of the variation in first-trimester size estimates.4PubMed. Impact of ovulation and implantation timing on first-trimester crown-rump length and gestational age This means a baby that implanted a day or two late might measure a few days “behind” on the first scan without there being any problem at all.
Later in pregnancy, a gap between GA and AUA is more likely to reflect the baby’s actual growth pattern rather than a dating error. A baby measuring ahead may be constitutionally large, or the mother may have undiagnosed gestational diabetes. A baby measuring behind could be constitutionally small, or it could signal that the placenta is not delivering enough nutrients. The clinical task shifts from “is the date right?” to “is the baby healthy?”
When Your Due Date Gets Changed
Providers do not adjust your due date every time the ultrasound age is slightly different from the LMP age. Guidelines from the American College of Obstetricians and Gynecologists (ACOG) recommend establishing the estimated due date (EDD) as early as possible and reserving changes for rare circumstances, documenting any revision clearly.5PubMed. Methods for Estimating the Due Date ACOG also considers a pregnancy without an ultrasound before 22 weeks that confirms or revises the EDD to be “suboptimally dated,” which can affect how decisions about induction or preterm delivery timing are made.
The thresholds for redating depend on how far along you are when the scan happens. In the first trimester, a discrepancy of more than about five to seven days between LMP and ultrasound dating usually triggers a revision to the ultrasound date. In the second trimester, the window widens to roughly ten to fourteen days. By the third trimester, ultrasound dating is much less precise, and providers rarely change the EDD based on a late scan unless there was no earlier ultrasound at all. When menstrual dates are unknown and the first scan happens late, gestational age estimation still has reasonable accuracy up to about 26 weeks but grows progressively less reliable after that.6PubMed Central. Ultrasound‐based gestational‐age estimation in late pregnancy
Once a due date is locked in, it stays. Your 36-week scan might show the baby measuring at 38 weeks, but your provider will not push the due date forward. Instead, they will interpret the discrepancy as the baby being large for its gestational age and investigate possible causes.
Measuring Ahead or Behind and What It Means
When AUA runs ahead of GA by a meaningful margin later in pregnancy, the baby is described as “large for gestational age” (LGA). This can be completely normal, especially if both parents are tall or if the mother has had large babies before. But it can also be a signal worth investigating. A large study found that fetuses identified as LGA on the second-trimester anatomy scan had roughly two and a half times the odds of the mother developing gestational diabetes and nearly seven times the odds of the baby being LGA at birth, even after adjusting for factors like obesity and maternal age.7PubMed Central. Large-for-gestational age diagnosed during second-trimester anatomy ultrasound and association with gestational diabetes and large-for-gestational age at birth That does not mean every big baby on ultrasound leads to a diabetes diagnosis, but it does mean your provider may recommend a glucose screening if one has not been done.
When AUA falls behind GA, the baby may be described as “small for gestational age” (SGA), generally defined as a birth weight below the tenth percentile for gestational age. Accurate dating is critical here because it determines whether you are looking at a truly small baby or one whose gestational age was overestimated. A baby that seems small at 34 weeks but is actually only 32 weeks is not small at all; the dates are just wrong.8PubMed Central. Small for gestational age: Case definition & guidelines for data collection, analysis, and presentation of maternal immunisation safety data
If the dates are right and the baby is genuinely small, the question becomes whether this is a constitutionally small but healthy baby or one experiencing fetal growth restriction, where the placenta is not functioning well enough. Clinical clues like reduced amniotic fluid or abnormal blood flow in the umbilical artery help providers make that distinction.6PubMed Central. Ultrasound‐based gestational‐age estimation in late pregnancy Umbilical artery Doppler, which measures the resistance of blood flow through the umbilical cord, is a key tool for assessing how severe the restriction may be and for deciding when delivery is safer than continuing the pregnancy.9Lynchburg Journal of Medical Science. Umbilical Artery Doppler in Fetal Growth Restriction
One nuance that often gets overlooked: standard growth charts compare your baby to a general population average, but babies naturally come in different sizes depending on the mother’s height, weight, ethnicity, and whether she has had babies before. Customized growth curves that account for these factors can reduce false alarms. Research has shown that applying customized standards, rather than a one-size-fits-all chart, lowers the number of babies incorrectly flagged as growth-restricted.10PubMed. Reduction of false-positive diagnosis of fetal growth restriction by application of customized fetal growth standards
How Accuracy Changes Across Trimesters
The single most important thing about pregnancy dating is that earlier is better. In the first trimester, CRL measurements can pin down gestational age to within a few days. One study using IVF pregnancies, where the exact date of embryo transfer is known, found that first-trimester dating ultrasound was accurate to within roughly a day and a half of the true gestational age.11PubMed Central. Validation of gestational age determination from ultrasound or a metabolic gestational age algorithm using exact date of conception in a cohort of newborns conceived using assisted reproduction technologies That is remarkably good.
By the second trimester, the margin of error grows to plus or minus one to two weeks. By the third trimester, it can be three weeks or more in either direction. This is not because the ultrasound machine is less precise later; it is because babies start growing at more individual rates as pregnancy progresses. A baby whose femur is a few millimeters longer than average at 32 weeks could be perfectly on track or could be a week “ahead,” and the ultrasound cannot tell the difference. That widening uncertainty is precisely why providers lock in the due date early.
International standards, such as those from the INTERGROWTH-21st project, were developed by tracking fetal measurements every five weeks from 14 weeks to birth in a carefully selected low-risk cohort, producing growth curves meant to apply across populations. These charts require that gestational age be confirmed by a first-trimester CRL measurement before the fetus enters the study, illustrating how foundational early dating is to everything that follows.12PubMed. International standards for fetal growth based on serial ultrasound measurements: the Fetal Growth Longitudinal Study of the INTERGROWTH-21st Project
IVF and Other Assisted Reproduction Pregnancies
If you conceived through IVF, you have something most pregnancies lack: a known date of embryo transfer, which gives you an almost exact conception date. You might expect the ultrasound to match this precisely, but studies consistently show a small systematic overestimation. One study found that ultrasound overestimated gestational age by an average of about three days compared to the embryo transfer date, using several different CRL reference charts.13PubMed Central. Assessing the accuracy of ultrasound estimation of gestational age during routine antenatal care in in vitro fertilization (IVF) pregnancies A separate study found a smaller but still statistically significant overestimation, averaging just under a day.14Medicinski pregled. IVF pregnancy dating: Comparative analysis of gestational age estimation based on embryo transfer date and ultrasound measurements of crown-rump length
This built-in overestimation is likely baked into the reference charts themselves, which were developed from populations where exact conception dates were not known. The charts may embed a slight forward bias because some of the women used to build them ovulated a day or two earlier than assumed. For IVF pregnancies, this means the embryo transfer date is generally considered the gold standard for dating, with ultrasound serving as a confirmation rather than a correction. A discrepancy of a few days in an IVF pregnancy is expected and usually not a cause for concern.
What Serial Scans Show That a Single Scan Cannot
A single ultrasound gives you a snapshot; it tells you how big the baby is today. Two or more ultrasounds spaced a few weeks apart give you a growth trajectory, which is far more informative. A baby that measures at the 15th percentile on both scans is tracking along its own curve and is likely fine. A baby that drops from the 50th percentile to the 15th percentile over four weeks is slowing down, and that trend matters far more than either measurement alone.
This is why providers do not simply compare AUA to GA at every visit and panic when the numbers shift. They are looking at how the measurements change over time. A fetus whose growth rate falls into the below-average range on serial third-trimester scans is at higher risk for being small at birth, even when a single scan might not have raised concern. Following the trajectory is how clinicians catch problems early enough to intervene, whether that means increased monitoring, steroid injections to mature the baby’s lungs, or delivery planning.
Artificial Intelligence in Ultrasound Dating
A newer development in gestational dating involves artificial intelligence. AI models have been trained to detect standard ultrasound planes, identify the right anatomical structures, and measure them automatically, reducing the operator-dependent variability that can creep into manual measurements.15PubMed Central. Application and Progress of Artificial Intelligence in Fetal Ultrasound One system, called DeepGA, was evaluated on over 10,000 ultrasound images and achieved a mean estimation error of about five days, outperforming the traditional manual measurement approach on the same dataset.16PubMed. DeepGA for automatically estimating fetal gestational age through ultrasound imaging
These tools are not yet standard in every clinic, but they point toward a future where gestational age estimation is less dependent on the skill and experience of the individual sonographer. That could be especially meaningful in resource-limited settings where trained ultrasound technicians are scarce. For now, the core method remains the same: measure early, measure carefully, and interpret any gap between GA and AUA in the context of everything else known about the pregnancy.
Early CRL and Long-Term Outcomes
An interesting area of research connects very early CRL measurements not just to dating accuracy but to longer-term pregnancy outcomes. A study examining CRL measured at specific days during early pregnancy found that measurements taken around the 67th day (roughly nine and a half weeks) were significantly smaller in babies who went on to have low birth weight. A CRL of about 26.5 millimeters at that gestational age had the highest predictive power for identifying low-birth-weight outcomes.17PubMed Central. Crown-rump length measured in the early first trimester as a predictor of low birth weight This suggests that the very first ultrasound measurement of your baby may carry information about growth potential that extends well beyond confirming a due date, though it is far from definitive on its own and would never be used in isolation to predict problems.