A baby’s weight is one of the first pieces of information that parents can share with their friends and family. The average baby born at full term weighs around 7 lbs., 5 oz., or 3,400 grams (Donahue et al. 2010). If a baby is born much larger than average, they are considered a “big baby.” And if it’s thought before birth that the baby might be born weighing much larger than average, then they are considered a “suspected big baby.” The purpose of this Evidence Based Birth® Signature Article is to share the evidence on the accuracy of the “suspected big baby” label during pregnancy, and to describe the health conditions and interventions that are commonly associated with big babies.

What is a big baby?

Some researchers consider a baby to be big when it weighs 4,000 grams (8 lbs., 13 oz.) or more at birth, and others say a baby is big if it weighs 4,500 grams (9 lbs., 15 oz.) or more (Rouse et al. 1996). Babies are called “extremely large” if they are born weighing more than 5,000 grams (11 lbs.) (Hehir et al. 2015).

A baby is also called “large for gestational age” or LGA if its weight is greater than the 90th percentile at birth. In other words, LGA is when a baby is born weighing more than 90% of all other babies born at that same gestational age (Rouse et al. 1996).

The medical term for big baby is macrosomia, which literally means “big body.” A baby cannot be diagnosed with macrosomia until after birth, when it is weighed for the first time.

How common are big babies?

About one in ten babies in the United States (U.S.) is born big. Overall, 8.78% of all babies born at 39 weeks or later weigh between 8 lbs., 13 oz., and 9 lbs., 15 oz., and only 1.17% are born weighing 9 lbs., 15 oz. or more (U.S. Vital Statistics 2026). In Table 1, you can see the percentages listed separately babies born to people who are not diabetic, vs. babies born to those with gestational diabetes and Type I or Type II diabetes.

The rate of big babies has decreased over time. In a study of all recorded births in the U.S. between 1971 and 2017 (with more than 147 million babies), researchers found that the rate of macrosomia decreased from 8.84% to 8.07% (Salihu et al. 2020a). The authors noted that improved treatment for women with gestational diabetes and better fetal monitoring may be the reason for this decrease in macrosomia.

What are some of the factors linked with having a big baby?

Big babies run in families (this is influenced by genetics), and it’s more common to have a big baby when the baby’s sex is male (Araujo Júnior et al. 2017). People with diabetes before or during pregnancy have higher rates of big babies compared to people who are non-diabetic.

Among people with gestational diabetes, researchers have found that having a higher blood sugar at first diagnosis makes you more likely to have a baby who is large for gestational age. (Metzger et al. 2008). However, pregnant people who manage their gestational diabetes through diet, exercise, or medication can bring down their chances of having a big baby to typical levels (or about 7%) (Landon et al. 2009).

In addition, there is high-quality evidence from 15 randomized trials showing that pregnant parents who exercise (both those with and without diabetes) have a significant decrease in the chance of having a big baby compared to those who do not exercise during pregnancy (Davenport et al. 2018).

Researchers have found other factors that are linked to a higher chance of having a big baby (Araujo Jr. et al. 2017; Bair et al. 2024; Dai et al. 2019; Fang Fang et al. 2019; Nguyen and Ouzounian 2021) such as:

• Higher body mass index (BMI) before pregnancy
• Higher weight gain during pregnancy
• Pregnancy at an older age
• Post term pregnancy (going past 42 weeks 0 days)
• A prior history of having a big baby
• Certain genetic conditions (including Beckwith-Wiedemann syndrome, Sotos syndrome, and Weaver syndrome)

What kind of care is frequently offered for suspected big babies?

At the end of pregnancy, during the third trimester, your provider will monitor the baby’s growth by measuring your fundal height, or the distance between the top of the uterus (i.e. fundus) and the pubic bone. They are trying to assess if the baby is either too small or too large for their gestational age. In high-risk pregnancies, providers may perform a third trimester ultrasound to check on the baby’s growth and development. It is at this time that providers may begin to voice concerns that the baby is big (ACOG 2016, reaffirmed 2025). 

Some of the best evidence we have about typical care for big babies in the U.S. comes from the Listening to Mothers survey series. The most recent survey in this series was published in 2026 (Sakala et al. 2026).

While roughly one in ten babies will be born large, Sakala et al. found that one in four (25%) participants reported that their provider told them that their baby would be large (≥ 4,000 g). Of this group, 74% had babies born in the normal weight range (2,500 g to 4,000 g) and 4% had babies who were considered small (≤ 2,500 g). Only 22% of those who were predicted to have a large baby had a large baby, and only 6% had a baby who weighed more than 4,500 g (Sakala et al. 2026).

Among those who were not told their babies would be large, 84% had babies born in the normal weight range, 12% had babies born small, and 4% had babies born weighing more than 4,000 g.

Sakala et al. also identified how the suspicion of a big baby could impact how people gave birth. Of the 25% of their survey respondents whose providers predicted they would have a big baby, 40% reported that their provider recommended they should have a labor induction due to the expected size of the baby. And 28% of those predicted to have a big baby reported that their provider recommended they have a Cesarean because their baby was expected to be large.

The Listening to Mothers research series provides important insights into the reality of maternity care in the U.S. from a perspective that centers birthing people’s experiences and the importance of respectful maternity care. We recommend them as a resource for anyone interested in pregnancy and childbirth in the U.S. Access the latest study and previous editions here at https://nationalpartnership.org/health-justice/maternal-health/listening-to-mothers-iv/.

So, if the baby appears large, a care provider will usually recommend either an induction or an elective Cesarean. Whether or not this approach is evidence based depends on the answers to five major questions:

1. Do big babies have a higher chance of their shoulders getting stuck (also known as shoulder dystocia)?
2. Are big babies at higher risk for other birth problems?
3. Can we accurately tell if a baby will be big?
4. Does induction keep the baby from getting any bigger, or lower the risk of Cesarean?
5. What are the benefits and risks of elective Cesareans for big baby?

Question #1: Do big babies have a higher chance of their shoulders getting stuck (also known as shoulder dystocia)?

One of the main concerns providers have with big babies is a complication called shoulder dystocia (“dis toh shah”). Shoulder dystocia occurs when the baby’s head has been born but one or both shoulders are stuck in the pelvis. Shoulder dystocia requires the care provider to take extra physical action(s), or maneuvers, to help get the baby out (ACOG 2017, reaffirmed 2024).

A shoulder dystocia by itself is not considered a “bad outcome.” It’s only a bad outcome if an injury occurs as a result of the shoulder dystocia (Personal communication, Emilio Chavirez, MD, FACOG, FSMFM).  The fear with shoulder dystocia is that the baby being stuck could result in several different bad outcomes. If there is too long of a delay between the birth of the head and the rest of the baby, it is possible that the baby may experience low oxygen, which could lead to brain damage. In rare cases, the maneuvers performed to help remove the baby could result in the baby having a broken clavicle (collarbone). The clavicle could be broken by accident, or the provider may break it on purpose to get the baby out in an emergency (Lau et al. 2024).

There is also a rare risk that the baby will experience a temporary or permanent nerve injury that could impact their ability to use their shoulder, arm, and/or hand. This injury is called brachial plexus palsy (which we describe in more detail below).

How often does shoulder dystocia occur?

Shoulder dystocia occurs in 0.2% to 3% of all vaginal births of babies in a head-first position (ACOG 2017, reaffirmed 2024). The chance of experiencing shoulder dystocia increases with the birth weight of the baby (Gherman et al. 2006).

Diabetes can also significantly increase the chances of shoulder dystocia, due to the impact of maternal blood sugars on a baby’s size.

One high-quality study that described rates of shoulder dystocia in pregnant people with and without diabetes showed that in non-diabetic people, shoulder dystocia happened to 0.65% of babies who weighed less than 8 lbs., 13 oz. (6.5 cases out of 1,000 births), 6.7% of babies who weighed between 8 lbs., 13 oz. and 9 lbs., 15 oz. (60 out of 1,000), and 14.5% of babies who weighed 9 lbs., 15 oz. or greater (145 out of 1,000) (Rouse et al. 1996).

In contrast, rates of shoulder dystocia were much higher in big babies whose birthing parent had Type I or Type II diabetes (2.2% of babies that weighed less than 8 lbs., 13 oz., 13.9% of babies that weighed between 8 lb., 13 oz. and 9 lb., 15 oz., and 52.5% of babies that weighed more than 9 lb., 15 oz.) (Rouse et al. 1996).

We were not able to find exact numbers for the percentage of people with gestational diabetes who had a baby with shoulder dystocia, as the rates change depending on each person’s blood sugar level. However, there is strong evidence that treatment for gestational diabetes drastically lowers the chance of having a big baby and shoulder dystocia. We cover the evidence on treatment for gestational diabetes (link evidencebasedbirth.com/inducingGDM) in our Evidence Based Birth® Signature Article on Induction for Gestational Diabetes.

Interestingly, people with high blood sugar levels during pregnancy have higher chances of shoulder dystocia during birth, even if their baby is not big. This is because weight can be distributed differently on a baby when they are exposed to high blood sugars during pregnancy. It’s more likely that the baby’s head size will be relatively small compared to the size of its shoulders and abdomen, which can make the birth of the shoulders more difficult (Kallem et al. 2020; Kamana et al. 2015).

Although big babies are at higher risk for shoulder dystocia, at least half of all cases of shoulder dystocia happen in smaller or normal sized babies (Morrison et al. 1992; Nath et al. 2015). This is because overall, there are more small and normal size babies being born than big babies.

Unfortunately, researchers have found that it is impossible to predict exactly who will have shoulder dystocia and who will not (Foster et al. 2011). Over the years, researchers have found that risk factors for shoulder dystocia have had a less than 2% to 3% positive predictive value (Gherman et al. 2006; Youssefzadeh et al. 2023). This means that for every 100 people who have one of these risk factors (like suspected big baby, or previous shoulder dystocia), less than 2 or 3 will experience shoulder dystocia– the other 97 to 98 will not.

Because at least half of shoulder dystocia cases occur in babies that are not big (Langer et al. 1991), and we can’t predict with any certainty who will have a shoulder dystocia, shoulder dystocia will always be a possibility during childbirth. The risk of shoulder dystocia could only be eliminated if all babies were born by Cesarean. Because requiring everyone to have a Cesarean is unethical, impractical, and would have other negative health implications, it is important for health care providers to train for the possibility of treating a shoulder dystocia.

Brachial plexus injury

Although most cases of shoulder dystocia can be safely managed by a care provider during the birth, some cases can result in a nerve injury in the baby called brachial plexus injury.

Brachial plexus injury, sometimes called brachial plexus palsy, can lead to weakness or paralysis of the arm, shoulder, and/or hand. This type of birth injury happens in about 0.9 to 1.5 out of every 1,000 vaginal births in the U.S. (Lin and Samora 2022).

A baby does not have to have shoulder dystocia to experience a brachial plexus injury—in fact, as many as half of brachial plexus injury cases happen without shoulder dystocia (De Francisco et al. 2019; Gherman et al. 1999). In one study that looked at 387 children who experienced brachial plexus injury, 92% were born vaginally and 8% were born by Cesarean (Chang et al. 2016). Other researchers have found that brachial plexus injury happens in about 3 per 10,000 Cesarean births (Chauhan et al. 2014).

A 2019 study of all births in Canada between 2004 and 2012 found that the type of Cesarean birth can also play a role (Augustine et al. 2019). Reviewing data from over 2 million births, researchers found that compared to an elective Cesarean birth, an emergency Cesarean was a moderate risk factor for a brachial plexus injury. A 2023 systematic review of 39 studies of brachial plexus injury after Cesarean identified that 54% of these injuries were associated with risk factors (Jaufuraully et al. 2023). The risk factors for brachial plexus injury after Cesarean included fetal malpresentation, obstructed labor (when labor is stalled because the baby does not fit through the birth canal due to the baby’s size, position, or issues with the pelvis caused by disease or malnutrition) (Pavličev et al. 2020), failed use of forceps or vacuum, emergency Cesarean, large baby or high maternal BMI, and Zavanelli’s maneuver (a maneuver of last resort in cases of severe shoulder dystocia, when a baby is pushed back into the birth canal to prepare for an emergency Cesarean) (Jaufuraully et al. 2023).

Some infants who have a brachial plexus injury (about 10% to 18%) will end up with a permanent injury, defined as arm or shoulder weakness or paralysis that persists for more than a year after birth. It’s estimated that there are anywhere from 35,000 to 63,000 people living with permanent brachial plexus injuries in the U.S. (Chauhan et al. 2014). For a blog article about what it’s like to grow up with a brachial plexus injury, read Nicola’s story here.

In 2019, researchers combined data from five studies about brachial plexus injury in pregnancies with babies over 8 lbs., 13 oz. versus those with babies who were not big (Beta et al 2019). Big babies had significantly more cases of brachial plexus injury (0.74% versus 0.06%). When babies weighed more than 4,500 grams (9 lbs., 15 oz.), the rate went up to 1.9%, or 1 in 50.

In a 2015 study of infants who were all extremely large at birth (>5,000 grams, or >11 lbs.), 17 of 120 infants born vaginally had shoulder dystocia, and three of those 17 had temporary brachial plexus injury that healed within six months—for an overall rate of about 1 brachial plexus injury cases per 40 vaginally-born, extremely large babies (Hehir et al. 2015).

In 1996, Rouse et al. published rates of shoulder dystocia and brachial plexus injury by infant weight. Using the numbers of permanent disability published by Chauhan et al. in 2014, we created Table 3 to show the differences between the weight groups.

Importantly, research has shown that when health care workers take part in annual inter-professional training (this means doctors, nurses, and midwives training together as a team) on how to handle shoulder dystocia, they can lower—and in some cases eliminate—brachial plexus injury among babies who experience shoulder dystocia (Brogaard et al. 2022; Crofts et al. 2016). Doctors have been trying to take this successful training (called “PROMPT”) from the United Kingdom and implement it in the U.S. Results at the University of Kansas showed a decline and then an eventual elimination of permanent cases of brachial plexus injury with PROMPT annual trainings (Weiner et al. 2015). Another example of this type of training is Advanced Life Support in Obstetrics (ALSO), a training offered by the American Academy of Family Physicians. The ALSO training prepares providers to collaborate as a team to respond to obstetric emergencies, including shoulder dystocia. (https://www.aafp.org/cme-and-events/topics/maternity-reproductive-womens-health/also)

Can a baby die or experience brain damage from shoulder dystocia?

Deaths from shoulder dystocia are possible but rare. In 1996, researchers looked at all the studies so
far that reported the rate of death due to shoulder dystocia. In 15 studies, there were 1,100 cases
of shoulder dystocia and no deaths (a death rate of 0%). In two other studies, the rates of infant death were 1% (one baby out of 101 “died at delivery,” possibly due to the shoulder dystocia) and 2.5% (one infant died out of 40 cases of shoulder dystocia) (Rouse et al. 1996).

In a study published by Hoffman et al. in 2011, researchers looked at 132,098 people who gave birth at term to a baby in head- first position. About 1.5% of the babies had a shoulder dystocia (2,018 cases), and of those, 101 newborns were injured. Most of the injuries were brachial plexus injury or collarbone fractures. Out of the 101 injured infants, there were zero deaths and six cases of brain damage due to lack of oxygen. With the six brain-damaged infants, it took an average of 11 minutes between the birth of the head and the body.

Mendez-Figueroa et al. investigated negative outcomes related to shoulder dystocia. Their 2021 study included 130,058 singleton births (not twins or other multiples), living babies born after 34 weeks in the U.S. between 2002 and 2008. The researchers found that shoulder dystocia occurred in 1.7% of the births (2,159 cases). Out of these, there were 5 cases of neonatal asphyxiation (when the baby does not get enough oxygen before, during, or immediately after birth) and one case of hypoxic ischemic encephalopathy (a type of brain injury related to a lack of oxygen) (Mendez-Figueroa et al. 2021).

Question #2: Are big babies at higher risk for other birth problems?

The risk of health problems with a big baby can increase along a spectrum. Lower risks are at 8 lbs., 13 oz., then higher risks are present at 9 lbs., 15 oz. The highest risks of health problems are associated with birth weights of 11 pounds or greater. In addition, the care provider’s “suspicion” of a big baby carries its own set of risks.

Unplanned Cesareans

Researchers combined 10 studies (called a meta-analysis) and found that babies with birth weights over 4,000 grams (8 lbs., 13 oz.) were more likely to have labors that end in Cesarean (Beta et al. 2019). In these studies, the average Cesarean rate was 19.3% for big babies versus 11.2% for babies who were not big. When babies weighed more than 4,500 grams (9 lbs., 15 oz.), the Cesarean rate increased to 27%. As we will discuss later, a care provider’s “suspicion” of a big baby can impact their likelihood of recommending Cesarean during labor.

Perineal Tears

In the meta-analysis published by Beta et al. (2019), five studies found a significant increase in the odds of severe perineal tears (also known as 3^rd or 4^th degree tears) with big babies, while three studies did not find a difference. When the researchers combined the results from all eight studies, the overall result showed that those who give birth to big babies are more likely to have severe perineal tears. The risk of a severe tear was 1.7% with big babies versus 0.9% for babies who were not big. When babies weighed more than 4,500 grams (9 lbs., 15 oz.), the rate of severe tears was 3%.

The largest single study on this topic (over 350,000 pregnant participants from National Health Service hospitals) examined 3rd degree tears and found the rate to be 0.87% with big babies versus 0.45% without (Jolly et al. 2003). In this study, pregnancies with big babies were also more likely to have longer first and second stages of labor and more use of vacuum and forceps. The increase in the use of vacuum and forceps among big babies likely contributed to the increase in severe tears.

The second largest study on this topic, which included over 146,000 hospital births in California between 1995 and 1999, found a higher rate of 4th degree tears in big babies who were born vaginally (Stotland et al. 2004). However, 4th degree tear rates in this study were unusually high, even among normal weight babies (1.5%), and the authors did not describe how many birthing people had episiotomies (a surgical cut of the perineum with scissors), which is a leading cause of severe tears.

A 2020 study of planned homebirths looked at maternal and infant outcomes when babies were born large (Pillai et al. 2020). This study included 68,966 homebirths that took place between 2012 and 2018. The researchers grouped babies into average-sized and macrosomic, or big. The 55,378 babies who were average sized (2,500 to 3,999 grams) were compared to the 13,588 babies considered large. The researchers also separated big babies into three categories: grade 1 macrosomia (4,000 to 4,499 grams; making 16% of the babies), grade 2 macrosomia (4,500 to 4,999 grams; making 3.3% of the babies), and grade 3 macrosomia (≥ 5,000 grams; making up 0.4% of the babies in the study). Compared with average sized newborns, perineal tears were more likely with larger babies. Perineal tears occurred in 1.3% of births of average sized babies, 1.8% of grade 1 macrosomia, 2.1% of grade 2 macrosomia, and 2.4% of grade 3 macrosomia.

Although having a big baby may be a risk factor for severe tears, it may be helpful to compare this risk to other situations that can also increase the risk of tears. For example, one large study found that the risk of a severe tear with a big baby ranged from 0.2% to 0.6% (Weissmann-Brenner et al. 2012). Other researchers have found that a vacuum delivery increases the risk of a severe tear by 11 times. So, if your baseline risk was 0.2%, it would increase to 2.2% with a vacuum, and the use of forceps increases the risk of a severe tear by 39 times (from 0.2% to 7.8%) (Sheiner et al. 2005).

Postpartum Hemorrhage

Researchers combined nine studies that reported on postpartum hemorrhage in people who gave birth to big babies compared to those who birthed babies who were not big (Beta et al. 2019). They found a higher rate of hemorrhage with babies over 8 lbs., 13 oz. (4.7% versus 2.3%). When the birth weights were over 4,500 grams (9 lbs., 15 oz.), the rate of postpartum hemorrhage was 6%. However, it is not clear whether this higher rate of postpartum hemorrhage is due to the big babies themselves or the inductions and Cesareans that often go along with a suspected big baby (Fuchs et al. 2013)—as both these procedures can increase the risk of postpartum hemorrhage (Magann et al. 2005).

However, in the study of homebirth outcomes where elective inductions and planned Cesareans were not used (Pillai et al. 2020), large babies were also associated with a greater risk of postpartum hemorrhage. Postpartum hemorrhage occurred in 2.7% of births of average sized babies, 4.2% of grade 1 big babies, 5.2% of grade 2 big babies, and 10.3% of the largest babies.

Newborn complications

One study compared 2,766 large babies with the same number of babies born with normal birth weights. All babies in the study were born to non-diabetic parents (Linder et al 2014). The researchers found that big babies were more likely to have low blood sugar after birth (1.2% vs. 0.5%), temporary rapid breathing (also known as “transient tachypnea” or “wet lung,” 1.5% vs. 0.5%), high temperature (0.6% vs. 0.1%), and birth trauma (2% vs. 0.7%).

The researchers did not say whether care providers suspected that the babies were large before labor began, or if their care was managed differently. More of the large infants in this study were born by Cesarean (33% vs. 15%), which could play a role in the higher rates of breathing problems, since breathing problems are more common with Cesarean-born babies.

Birth fractures, or broken collarbones or arms, are rare but more likely to occur among big babies. Researchers combined the results from five studies and found that the rate of birth fractures among babies over 4,000 grams (8 lbs., 13 oz.) was 0.54% versus 0.08% among babies who are not big (Beta et al. 2019). When babies weighed more than 4,500 grams (9 lbs., 15 oz.), the fracture rate increased to 1.01%.

In their 2020 study about big babies and home births, Pillai et al. looked at a variety of health outcomes in relation to newborn birth weight. Big babies were more likely than normal weight babies to have a birth injury (0.6% vs. 0.2%), respiratory distress (2.4% vs. 1.6%), and NICU stays that lasted more than 24 hours (2.3% vs. 1.9%).

Stillbirth

Some doctors recommend Cesareans for suspected big babies because they believe there is a higher risk of stillbirth. There have been three large studies on this topic (see below). In general, researchers have found that the smallest babies have the highest risk of stillbirth, while large babies tend to have lower or average risks of stillbirth. However, the very biggest big babies (weighing 5,000 grams or more) tend to have slightly higher-than-average risks of stillbirth, partially due to higher rates of maternal diabetes in this group.

Harms due to the Care Provider’s Suspicion of a Big Baby

When a big baby is suspected, families are more likely to experience a change in how their care providers view and manage pregnancy, labor, and birth. This leads to a higher Cesarean rate and a higher rate of people inaccurately being told that labor is taking “too long” or the baby “doesn’t fit.”

In fact, research has consistently shown that the care provider’s perception that a baby is big can be more harmful than an actual big baby by itself.

Overall, a total of ten different studies from 1992 to 2024 have all shown that it is the suspicion of a big baby—not big babies themselves—that can influence providers and lead to higher induction rates, higher Cesarean rates, and higher diagnoses of stalled labor (Levine et al. 1992; Weeks et al. 1995; Parry et al. 2000; Weiner et al. 2002; Sadeh-Mestechkin et al. 2008; Blackwell et al. 2009; Melamed et al. 2010; Little et al. 2012; Peleg et al. 2015; Sgayer et al. 2024). Some examples of these studies are provided below.

It’s not surprising that physicians are more likely to turn to Cesarean in situations where a big baby is suspected. In one medical journal editorial, an obstetrician with a preference of Cesarean for big babies said, “Flagging up all cases of predicted fetal macrosomia is vitally important, so that the attendants in the labor suite will recommend Cesarean if there is any delay in cervical dilatation or arrest of head rotation or descent. Cesarean should also be the preferred option if an abnormal fetal heart tracing develops” (Campbell 2014).

There is more work that needs to be done on this topic. A 2016 Cochrane Review on the suspicion of big babies found no randomized trials that evaluated the overall benefits and drawbacks of fetal growth surveillance for suspected big baby (Culliney et al. 2016). Although big babies are at higher risk for some problems, the care provider’s perception that there is a big baby carries its own set of risks. This perception—whether it is true or false—changes the way the care provider behaves and how they talk to families about their ability to birth their baby, which, in turn, increases the chance of Cesarean.

Question #3: Can we accurately tell if a baby will be big?

Both physical exams and ultrasounds have limited accuracy when predicting whether a baby will be big at birth.

Time and time again, researchers have found that it is very difficult to predict a baby’s size before it is born. Although two out of three people giving birth in the U.S. receive an ultrasound at the end of pregnancy (Declercq et al. 2013) to “estimate the baby’s size,” both the care provider’s estimate of the baby’s size and ultrasound results can be unreliable. Two different studies even found that pregnant women’s own estimate that their baby was over 4,000 grams was as accurate as a health care provider who used clinical palpation (measuring the baby through touching the outside of the pregnant person’s abdomen) (ACOG 2020, reaffirmed 2026; Peregrine et al. 2007). 

After a baby is born, measuring their weight is as easy as placing them on the scale. However, figuring out how much something weighs when you can only feel it with your hands or take measurements using ultrasound is a different challenge. Ultrasound fetal weight estimates rely on mathematical formulas to make an educated guess on how much the fetus weighs, based on ultrasound measurements of different body parts. A number of factors can impact the accuracy of these fetal weight estimates, including operator experience or errors, the technical difficulty of taking certain measurements, the position of the baby, and the body size of the pregnant person (Milner and Arezina 2018; Saloman et al. 2019). If just one measurement is inaccurate, that inaccuracy can be multiplied when it is entered into a mathematical formula used to estimate fetal weight (Salomon et al. 2019).

In 2005, researchers looked at all the studies that had ever been done on ultrasound and estimating the baby’s weight at the end of pregnancy. They found 14 studies that looked at ultrasound and its ability to predict that a baby would weigh more than 8 lbs., 13 oz. Ultrasound was accurate 15% to 79% of the time, with most studies showing that the accuracy (“post-test probability”) was less than 50%. This means that for every ten babies that ultrasound predicts will weigh more than 8 lbs., 13 oz., five babies will weigh more than that and the other five will weigh less (Chauhan et al. 2005).

Ultrasound was even less accurate at predicting babies who will be born weighing 9 lbs., 15 oz. or greater. In three studies that were done, the accuracy of ultrasounds to predict extra-large babies was only 22% to 37%. This means that for every ten babies the ultrasound identified as weighing more than 9 lbs., 15 oz., only two to four babies weighed more than this amount at birth, while the other six to eight babies weighed less (Chauhan et al. 2005).

The researchers found three studies that looked at the ability of ultrasound to predict big babies in pregnant people with diabetes. The accuracy of these ultrasounds was 44% to 81%, which means that for every ten babies of a diabetic parent who are thought to weigh more than 8 lbs., 13 oz., around six will weigh more and four will weigh less. The ultrasound test probably performs better in diabetics simply because diabetics are more likely to have big babies. In other words, it’s easier to predict a big baby in someone who is much more likely to have a big baby to begin with.

Currently, there is no reason to believe that three-dimensional (3D) ultrasound is any better at predicting birth weight and big babies than two-dimensional (2D) ultrasound (Tuuli et al. 2016). Research is ongoing to determine if 3D measurements can be combined with 2D measurements to better predict macrosomia.

Overall Accuracy of Identifying Big Babies in Utero

Compared to using ultrasound, care providers are just as inaccurate when it comes to using a physical exam to estimate the size of the baby. However, ultrasound appears to provide more accurate estimates when pregnant people are plus size (Preyer et al. 2019).

Overall, when a care provider estimates that a baby is going to weigh more than 8 lbs., 13 oz., the accuracy is only 40-53% (Chauhan et al. 2005). This means that out of all the babies that are thought to weigh more than 8 lbs., 13 oz., half will weigh more than 8 lbs., 13 oz. and half will weigh less.

The care provider’s accuracy goes up if the pregnant person has diabetes or is post-term, again, probably because the chance of having a big baby is higher among these groups. Unfortunately, all the studies that looked at diabetes and the accuracy of ultrasound grouped people with gestational diabetes and those with Type I or Type II diabetes into the same groups, limiting our ability to interpret these results.

Even the “best” way to predict a big baby is going to have problems identifying actual big babies—most often overestimating the size of the baby. In a 2010 study by Rosati et al., researchers tested different math formulas to figure out an infant’s estimated weight with ultrasound. The best formula for predicting birth weight was the “Warsof2” formula, which is based solely on the baby’s abdominal measurement. The results of this formula came within ±15% of the baby’s actual weight in 98% of cases. As an example, if your baby’s actual weight was 8 lbs., 0 oz. (3,629 grams), the ultrasound could estimate the baby’s weight to be anywhere between 6 lbs., 13 oz. (3,090 grams) and 9 lbs., 3 oz. (4,450 grams).

A systematic review concluded that there is “no clear consensus with regard to the prenatal identification, prediction, and management of macrosomia.” The authors explained that the main problem with big babies is that it is very difficult to diagnose big babies before birth—it’s a diagnosis that can only be made after birth (Rossi et al. 2013). While researchers continue to look for new or improved ways to improve the accuracy of fetal size estimates, it remains an imperfect science.

Question #4: Does induction keep the baby from getting any bigger, or lower the risk of Cesarean?

There is conflicting evidence about whether induction for suspected big babies can improve health outcomes.

We will talk about a few main pieces of evidence in this section:

• The 2025 “Big Baby” Trial from the UK
• A 2023 Cochrane review (when researchers combined multiple randomized trials together)
• The largest randomized trial from the Cochrane review (published in 2015)
• The second-largest study from the Cochrane review (published in 1997)

The Big Baby Trial (2025)

The “Big Baby Trial” (Gardosi et al. 2025) is the largest randomized controlled trial to test whether early labor induction for a suspected big baby could prevent shoulder dystocia. There were two main purposes of this study: 1) to determine whether early induction of labor is associated with a lower risk of shoulder dystocia compared with standard care, and 2) to investigate if waiting to induce until 38 weeks instead of 37 weeks for shoulder dystocia would still reduce risk.

The researchers recruited women with suspected big baby (estimated fetal weight was greater than something they called the “90th customized percentile”) between 35 to 38 weeks’ gestation from 106 hospitals across the United Kingdom (U.K.) between 2018-2022. People excluded from taking part in this trial included: those with Type 1 or 2 diabetes or gestational diabetes, those already had a scheduled induction or scheduled Cesarean, twins or multiples, and those whose babies were not in a head down position.

One concern at the time of trial design was that individuals who were otherwise eligible would, after being informed of the risks of shoulder dystocia and induction of labor, choose to have an elective Caesarean or would decline to be randomly assigned. Those who declined random assignment were invited to join an observational (non-randomized) study that would provide additional data on outcomes (as of the time of writing in 2026, this parallel study is not yet published).

Gardosi et al. estimated that an earlier birth would reduce the risk of shoulder dystocia by half. The researchers used data from a recent study in Europe (Boulvain et al. 2015), in which the rate of shoulder dystocia was 3.9%, to estimate they would see an overall shoulder dystocia rate of 4% rate in their trial. So, they estimated that in order to see a statistically significant decrease in that number by half with early induction, they would need about 4,000 participants.

However, recruitment was stopped early on the advice of the data monitoring committee due to an ongoing review of data showing that, in reality, more than 12,000 participants would be needed to see any significant results. This was because shoulder dystocia was so uncommon in this study, ultimately occurring just 38 times out of 2,893 total births (insert percentage here). The researchers had already been recruiting participants for nearly four years, and it would take a very long time to recruit more than 9,000 additional participants. At the time the study was stopped, the recruitment rate was only about 10 women per week. 

The researchers ultimately enrolled 2,893 participants. Participants were randomly assigned to one of two groups:

1. Induction of labor between 38 weeks 0 days and 38 weeks 4 days of pregnancy (1,447 participants)
2. Standard care, defined as the hospital’s usual care. In the standard care group, to help create enough of a difference in gestational age between groups, induction before 39 weeks 4 days was discouraged, unless there was a medical reason (1,446 participants)

The researchers analyzed results using two statistical methods (Smith et al. 2021):

• Intention to Treat (ITT), which keeps everyone in their original assigned group, including those who did not follow the randomly assigned group protocol. Intention to Treat is the preferred method for most trials, more representative of reality, and less prone to bias. It is also the gold standard for reporting data from randomized trials.
• Per Protocol (PP) only analyzes those who adhered to protocol for the randomly assigned group. Per Protocol is more prone to bias, but it can give a better sense of treatment efficacy in a perfect condition with strict protocol adherence. Data analysis with PP is rarely done when researchers publish results from randomized trials.

In the Intention to Treat analysis, the incidence of shoulder dystocia was 2.3% in the induction group and 3.1% in the standard care group. This difference was not statistically significant, meaning any difference between groups is most likely due to chance.

However, perhaps because the Intention to Treat analysis was not significant, the authors also published a Per Protocol analysis with 77.9% of the total participants: 81.5% of the induction group and 74.3% of the standard care group. The rate of shoulder dystocia in the Per Protocol induction group was 2.3% vs. 3.7% for the Per Protocol standard care group. Removing those participants who did not adhere to the study protocol made this a statistically significant difference and allowed the researchers to describe their trial as showing a decrease in shoulder dystocia with early labor induction.

Participants in the induction group were more likely to have a longer hospital stay (3.4 days vs. 3 days) and more likely to have a second-degree tear (28% vs 24.5%). They were less likely to have an elective Cesarean (2.7% vs 4.2%), less likely to have an emergency Cesarean (25.7% vs 28.8%), and less likely to have an instrumental birth (16.3% vs 17.9%), and less likely to have an episiotomy (19.8% vs 22.6%).

There were no significant differences between groups in other maternal health outcomes, including:

• 3rd or 4th degree tears (rates were approximately 46% to 50%)
• Adverse maternal outcomes or complications before hospital discharge (rates were about 6% to 7.5%)

There were also no significant differences between groups in newborn health outcomes, including:

• Admission to NICU (rates were about 9.6% to 10.7%)
• Birth injury (rates were about 0.1% to 0.3%)
• Complications from being born too early (rates were about 5.3% to 6.2%)

Two infant deaths occurred during the study. One baby died in the induction group as a result of sepsis (blood infection) that occurred with a prolonged induction. One baby died in the standard care group from perinatal asphyxia (lack of oxygen during birth) after shoulder dystocia.

So, what are the take aways from the Big Baby Trial? Early induction (prior to 39 weeks) for a suspected big baby might reduce the incidence of shoulder dystocia, but this result was only found when researchers used a non-gold-standard method of data analysis. Researchers also found that inductions do not need to take place at 37 weeks to see benefit, and that providers can wait until 38 weeks. In their discussion, Gardosi et al. recommended a policy of shared decision making that includes information about potential benefits to a 38-week induction, as well as limitations when it comes to estimating fetal weight via ultrasound.

In addition, any shared decision-making conversation that includes data from the Big Baby Trial should disclose that the researchers had to stop the study early because shoulder dystocia was so uncommon (2% to 3% among suspected big babies), and that they ultimately found no difference in NICU admission rates or birth injury rates. 

Cochrane Review on Induction at 37 to 40 weeks for Suspected Big Babies (2023)

In a 2023 Cochrane review, researchers combined four studies in which 1,190 non-diabetic pregnant people with suspected big babies were randomly assigned to either 1) induction between 37 and 40 weeks or 2) waiting for spontaneous labor (Boulvain et al. 2023).

When researchers compared the induction group to the waiting group, they found a decrease in the rate of shoulder dystocia in the induction group—about 41 cases per 1,000 births in the elective induction group, down from 68 cases per 1,000 in the waiting group.

They also found a decrease in birth fractures in the elective induction group (4 per 1,000 vs. 20 per 1,000 in the waiting group). To prevent one fracture, it would be necessary to electively induce labor in 60 people. The researchers found an increase in the treatment of newborn jaundice (11% vs. 7%) in the induction group.

On average, babies weighed 178 grams (6 ounces) less when labor was electively induced early, compared with those assigned to wait for labor.

There were no differences between groups in rates of Cesarean, vacuum- or forceps-assisted births, NICU admissions, brachial plexus injury, or low Apgar scores. Three of the four studies reported death rates, and there were zero deaths in either group.

Researchers did not look at patients’ satisfaction with their care or any long-term health results for birthing people or babies.

Question #5: What are the benefits and risks of elective Cesareans for big baby?

Although some care providers will recommend an induction for a suspected big baby, others skip this step and go straight to recommending an elective Cesarean. However, researchers have estimated that this type of approach is extremely expensive and that it would take thousands of unnecessary Cesareans to prevent one case of permanent brachial plexus injury.

In 1996, an important analysis published in the Journal of the American Medical Association proposed that a policy of elective Cesareans for all suspected big babies was not cost-effective and that there were more potential harms than potential benefits (Rouse et al. 1996).

In this analysis, the researchers calculated the potential effects of three different types of policies:

• No routine ultrasounds to estimate the babies’ sizes
• Routine ultrasounds, then elective Cesarean for babies estimated to weigh 8 lbs., 13 oz. or more
• Routine ultrasounds, then elective Cesarean for babies estimated to weigh 9 lbs., 15 oz. or more

The researchers looked at the results separately for diabetic and non-diabetic people. Unfortunately, most research up to this time point did not distinguish between Type 1 or Type II diabetes and gestational diabetes. So, the term “diabetic” could refer to any of these conditions.

Among non-diabetics, a policy of elective Cesarean for all suspected big babies over 8 lbs., 13 oz. means that a large number of pregnant people and babies would experience unnecessary surgeries. In order to prevent one permanent brachial plexus injury in babies suspected to be over 8 lbs., 13 oz., there would need to be 2,345 unnecessary Cesareans at a cost of $4.9 million dollars per injury prevented. The costs in this study were estimated using year 1995 dollars; adjusting for inflation, this would cost $10.9 million dollars per injury prevented in the year 2026.

With a policy of elective Cesareans for all suspected big babies over 9 lbs., 15 oz., even more pregnant people would have surgeries found to be unnecessary in retrospect, because ultrasounds are even less accurate in higher suspected weight ranges (Chauhan et al. 2005). In order to prevent one permanent brachial plexus injury in babies suspected to be over 9 lbs., 15 oz., 3,695 people would need to undergo unnecessary Cesareans at a cost of $8.7 million per injury prevented. Adjusting for inflation, this would cost $19.4 million dollars per injury prevented in the year 2026.

Policies encouraging elective Cesareans for suspected big babies could also potentially increase rates of known risks from Cesarean, such as serious infections, blood clot disorders, postpartum hemorrhage requiring blood transfusions, and newborn breathing problems.

Among diabetics, the results of the Rouse et al. study were different—mostly because ultrasound is slightly more reliable at predicting big babies in pregnant people who are diabetic, and because shoulder dystocia is more common
in this group as well. If pregnant diabetics were offered an elective Cesarean for every baby suspected of weighing more than 8 lbs., 13 oz., it would take 489 unnecessary surgeries to prevent
one case of permanent nerve damage, at a cost of $930,000 per injury avoided. Adjusting for inflation, this would cost $2 million dollars per injury prevented in the year 2026. If diabetics had elective Cesareans when their babies were suspected of being 9 lbs., 15 oz. or greater, it would take 443 unnecessary surgeries to prevent one case of permanent brachial plexus injury, at a cost of $880,000 per injury avoided. Adjusting for inflation, this would cost $1.9 million dollars per injury prevented.

Importantly, a cost-effectiveness analysis is only as good as its assumptions–the numbers that they use to plug into the analysis. For example, we might ask, how did Rouse et al. determine how frequently shoulder dystocia occurs, the accuracy of ultrasounds, and how many permanent injuries occur with a shoulder dystocia? In the Rouse et al. (1996) paper, the authors did a very high-quality literature review to determine these factors. One drawback of this analysis is that the costs they reported did not consider the cost of lawsuits due to birth injuries.

Another important drawback is that this analysis is now 30 years old.

Since the landmark Rouse et al. paper was published, two newer cost-effectiveness analyses have been published. However, both of these newer papers had major problems—one of them did not take into account the inaccuracy of ultrasound (Herbst 2005), and the other study had a poor-quality systematic review—using numbers in their assumptions that overestimate the accuracy of ultrasound (Culligan et al. 2005). Because these groups of researchers did not do a good job of making their assumptions, we cannot trust the results of their analyses, and so their results are not included in this Signature Article.

In summary, evidence does not support elective Cesareans for all suspected big babies, especially among non-diabetic pregnant people. There have been no randomized, controlled trials testing this intervention for big babies. And there are no high-quality research studies to see what happens when this intervention is used on a mass-scale in real life.

In fact, pregnant people without diabetes may be given one-sided information by their care providers if elective Cesarean is presented as a completely “safe” or “safer” option than vaginal birth for a suspected big baby. Although giving birth vaginally to a big baby carries risks, Cesarean surgery also carries potential harms for the birthing person, infant, and any children born in future pregnancies. It is important to have full and accurate information of potential benefits and risks of both options in order to participate in informed decision making.  

Guidelines

U.S. Guidelines:

In 2020, the American College of Obstetricians and Gynecologists (ACOG) released practice guidelines for big babies, which were reaffirmed in 2026. In these guidelines, ACOG states that estimating birthweight via ultrasound is an imprecise form of measurement, with accuracy no better than with measurement by hand (Level A evidence, recommendations are based on good and consistent scientific evidence). They also state that suspected macrosomia (big baby) is not an indication for induction of labor before 39 weeks of gestation because there is insufficient evidence that benefits of reducing shoulder dystocia risk would outweigh the harms of early delivery (Level B evidence, recommendations are based on limited or inconsistent scientific evidence). Their guidelines recommend that planned Cesarean may be considered for babies suspected to weigh more than 11 lbs. (5,000 grams) in women without diabetes, and 9 lbs., 15 oz. (4,500 grams) in women with diabetes and gestational diabetes (Level C evidence, based primarily on consensus and expert opinion). They also recommend that providers should counsel pregnant people about the risks and benefits of induction or Cesarean for suspected big baby (Level C evidence, based primarily on consensus and expert opinion).

Canada Guidelines:

In 2023 the Society of Obstetricians and Gynaecologists of Canada (SOGC) released guidance stating that by itself, suspected macrosomia (big baby) or suspected large for gestational age fetus is not an indication for induction of labor before 39 weeks (Conditional recommendation, low-quality evidence) (Robinson et al. 2023).

U.K. Guidelines

In 2021, the National Institute for Health and Care Excellence (NICE) released updated guidance on inducing labor, including recommendations for suspected big baby. The guidelines for someone without diabetes state that in the case of suspected macrosomia (>95% percentile estimated fetal weight), Cesarean, induction and expectant management are all options. They note that there is uncertainty about the benefits and risks of induction of labor compared to expectant management, but that there is reduced risk of shoulder dystocia with induction, an increased risk of third- or fourth-degree perineal tears with induction, and that the risk of perinatal death, brachial plexus injuries in the baby, or the need for emergency Cesarean birth do not differ between the 2 options. They recommend providing counseling about birth options, taking into account individual circumstances and preferences, and to respect the patient’s decision.

Australia and New Zealand Guidelines

In 2021, the Royal Australian and New Zealand College of Obstetricians and Gynaecologists (RANZCOG) released their latest guidance on suspected big babies. They note that ultrasound can be used to rule out a big baby, but that it is an imperfect predictor of shoulder dystocia and has limited ability to predict fetal weight (Level A evidence, body of evidence can be trusted to guide practice). The benefit of early induction (before 39 weeks) in babies predicted to have a >95% percentile estimated fetal weight must be weighed against limitations to using ultrasound to predict estimated fetal weight and potential complications from early term birth (Level B evidence, body of evidence can be trusted to guide practice in most situations). An elective Cesarean may be beneficial for newborns with suspected macrosomia who have an estimated fetal weight of 5,000 grams or more in women without diabetes, and with an estimated fetal weight of 4,500 grams or more in women with diabetes (Consensus based recommendation, based on clinical opinion and expertise as insufficient evidence is available).