Nub theory guesses the sex from the angle of the genital tubercle, a small bump between the legs, on a side view at about 12 weeks: pointing up suggests a boy, flat or down a girl. At a London fetal medicine center in 1999, researchers judged the angle while scanning 172 pregnancies at 11 to 14 weeks: 70.3% of calls at 11 weeks were right, and 98.7% at 12. A photo read later adds room for error, and some paid services and apps claim 99% or more without saying how they counted. No scan is 100% sure.
Inside: how the angle is read, what studies found, why photos mislead, and what works instead.
What is the “nub”?
The “nub” is the genital tubercle, a small bump of tissue between the baby’s legs. Boys and girls both have one, because the genitals of both sexes “come from the same tissue” (MedlinePlus). With male hormones (androgens), the tubercle grows into a penis. Without them, it becomes a clitoris (Baskin and colleagues, 2020).
At first, the two look alike. A University of California, San Francisco team calls boys’ and girls’ tubercles “indistinguishable” at 8 to 9 weeks after fertilization (Baskin and colleagues). That’s about 10 to 11 weeks counted from the last period (Mayo Clinic).
By about 12 weeks, a sonographer may be able to see which way it points. A tubercle pointing “up toward the baby’s head” suggests a boy, and one that “remains flat or points down” suggests a girl, an obstetrician at UT Southwestern Medical Center in Dallas explains (UT Southwestern, 2025). A scan detects the baby’s sex from the anatomy it shows. Can you find out the gender at 12 weeks? covers what has formed by then.
Where nub theory comes from
In 1989, radiologists at the University of Tennessee in Memphis described the “sagittal sign,” a way to predict the sex from a side view. They made 184 scans of 165 pregnancies at 10 to 20.5 weeks and checked predictions in the 105 whose sex was known from amniocentesis or chorionic villus sampling (CVS). At 14 to 20.5 weeks, the side view did better than the usual views (Emerson and colleagues, 1989). A 2009 review defines the sign as “the direction of the genital tubercle,” with downward suggesting a girl and upward a boy (Odeh and colleagues, 2009).
The angle came a decade later. In 1999, researchers at King’s College Hospital’s fetal medicine center in London scanned 172 single pregnancies at 11 to 14 weeks, just before CVS, a test of the baby’s chromosomes. They called a boy or a girl from the tubercle’s angle (Efrat and colleagues, 1999).
The aim was medical. In pregnancies at risk of a sex-linked genetic condition, invasive testing “would be necessary only in pregnancies with male fetuses,” the authors wrote. They suggested deciding on invasive testing only after 12 weeks. Later teams tested the angle in larger groups: 656 pregnancies in Israel in 2006, and several hundred at a hospital in Dayton, Ohio, in 2023. Today the method is sometimes called “nub theory,” as the UT Southwestern obstetrician notes.
How the angle is read
Studies describe these steps:
Get a true side view. The scan has to cut straight down the middle of the baby, from head to bottom. This is the midsagittal plane.
Wait for a neutral position. The baby should be neither curled up nor arched back. Otherwise “the angle may be overestimated or underestimated and the result may be misleading,” Spanish researchers warn, reporting on 2,314 first-trimester scans (Savirón-Cornudella and colleagues, 2015).
Draw a reference line. In the London method, it runs flat through the skin of the lower back, the lumbosacral skin. The Spanish team measured against the spine.
Measure the tilt. The angle between that line and the nub decides the call.
The cut-offs weren’t the same in every study:
London, 1999. Over 30 degrees, a boy. Flat along the line or angled toward it (under 30 degrees), a girl. The rule had no middle ground.
Spain, 2015. The same 30 and 10 degrees, against the spine. In-between angles, or a baby in an unsuitable position, got no call.
São Paulo, 2005. Two other angles of the tubercle, with cut-offs of 42 and 24 degrees (Bunduki and colleagues, 2005).
That middle zone builds “not sure yet” into the method: an in-between angle means no call, not a guess.
How accurate is nub theory?
What the studies found
In the London study, accuracy rose from 70.3% at 11 weeks to 98.7% at 12 and 100% at 13. The abstract gives no count for each week, and doesn’t say whether any baby went without a call. At 11 weeks, more than half the boys (56%) were called girls, while only 5% of girls were called boys. In boys, the angle grew as the baby grew.
Before 12 weeks, experts advise against trying. The Australasian Sonographers Association (ASA) said in 2022 that sex determination by ultrasound before 12 weeks “should be discouraged” (ASA). The 2009 review agrees: that early, “it is relatively inaccurate.”
Why accuracy figures differ so much
Figures differ for three reasons:
The week. The low figures come from the earliest scans, like the London study’s 11-week results above.
Who is counted. Leave out the babies who got no call, and a figure looks better: 91% instead of 75% in the Australian scans.
Who checked. The Australian practice checked each call against the baby’s sex after birth. One paid service we looked at in October 2026 asks customers to report back “to keep a tally for accuracy.”
Why a photo guess isn’t the same test
A guess from a scan picture skips what the method needs:
The right view. The studies above read side views, and several required an exact view down the middle. At the University of São Paulo, each photo had to show a strict side view, with the baby filling at least 75% of the screen. One operator, who didn’t know the sex, later measured the angles on the photos with a ruler and protractor. When a 2024 Saudi team reviewed two hospitals’ saved scans from 11 weeks to 13 weeks 6 days, about two in three cases didn’t have the required side view. It also left out 12 of 324 cases (3.7%) with a poor view of the genital area (Alfuraih and colleagues, 2024).
The right moment. The angle “can change during ultrasonographic examination,” already at 11 weeks, a Rotterdam team notes, citing a 2001 report (Bogers and colleagues, 2018). A still photo catches one moment.
Two careful pictures can disagree. The São Paulo study fully evaluated 455 of 508 cases at 11 weeks to 13 weeks 6 days. For the last 30, two operators each took their own side-view picture of the same baby, without seeing the other’s, and the angles were measured independently. On average, they differed by 9 and 15.7 degrees, for the two angles measured.
An app reads the same single picture, and in online groups, strangers guess from it. We found no published study testing nub guesses made by apps, paid services or groups.
Paid nub readings and apps: what they say
A 2018 news report described a UK parenting website’s survey of 1,275 parents. In it, 21% of mothers-to-be had posted their picture “for an expert to predict the sex,” against 7% who showed it to their midwife (National News, 2018).
In October 2026, we looked at four online nub-reading services, all priced in pounds, and three iPhone apps:
What they charge. About £10 to £25 for a nub reading, with results in as little as 15 minutes, one says. The apps are free to download, with purchases or weekly plans from $5.99.
What they claim. One service says it has a “99.2% accuracy rate,” and another’s page description for search engines says “99.9% accurate.” One app says its team “has achieved a remarkable 99% accuracy rate.” None says how its figure was counted.
Who reads. One service says its reader is self-trained, with over 16 years of experience. Another cites 15 years of experience “via social media groups, readings given on here and working in a private gender clinic.” The apps say they use AI.
What they don’t promise. All four services say they aren’t medical professionals or don’t give medical advice. Two apps call themselves “for entertainment purposes only.” Three services say they refund you if they can’t make a call or your picture is unsuitable, but not if the guess turns out wrong.
What else they sell. Three of the four also sell Ramzi readings, and one covers skull theory. Ramzi theory and skull theory look at the evidence for those.
Before you pay, ask:
How many readings is the accuracy figure based on, and at which weeks?
How was each guess checked: against the birth, a later scan, or customers who chose to report back?
What happens if the guess is wrong?
The bottom line, and what works
Nub theory isn’t made up. It borrows a method specialists have tested since the late 1990s, reading the angle on a true side view. A guess from a photo is a different test, and the services selling it say they don’t give medical advice. Enjoy it as a game, like our old wives’ tales quiz, but wait for a surer answer before you paint or plan a reveal.
These routes are surer, though none is 100% sure:
A blood test, from about 10 weeks. In the U.S., your doctor or midwife can order cell-free DNA screening (NIPT), which looks for chromosomal conditions. Because it checks the sex chromosomes too, it can report the sex as a side result. Cleveland Clinic calls that result “highly accurate, but not 100% accurate” (Cleveland Clinic).
A later scan. In the studies above, calls grew more accurate as the weeks passed.
The anatomy scan. Cleveland Clinic puts this medical scan at about 18 to 22 weeks in the U.S., and says you can learn the sex there “in most cases,” if the baby’s position allows (Cleveland Clinic). In England, it depends on the hospital’s policy, and a sonographer can’t be “100% certain,” the NHS says (NHS).
4D Bonding16 to 36 weeks. Faces look best from 28 to 34.Book 4D Bonding
Your date stays on this device. We never see it.
Still short of 13 weeks? Enter your due date, and the planner lists the sessions ready for you now and what opens after that. The date stays on your device.
Why we start at 13 weeks, in Glendale
At our studio in Glendale, Los Angeles County, we book from 13 weeks, and we’re surest from 15. If you can wait until 15, do: we’re more confident, and a second look is less likely. At 12 weeks? Sometimes, but rarely with confidence, so we book from 13 weeks.
Our Gender Confirmation is a 2D scan, “the classic black-and-white picture.” Bring whoever you like. Everyone in the room sees your baby on the big screen. When we can’t be sure, we say so, and you come back for a free re-check (our promise). Had a nub guess already? If what we see is different from what you’ve been told, we tell you what we see, and how sure we are. What a blood test result means is a question for your doctor or midwife. No ultrasound can promise 100%.
Results vary by study. In a 1999 London study of 172 pregnancies at 11 to 14 weeks, where researchers judged the angle during the scan, 98.7% of calls at 12 weeks were right, though the abstract gives no count for that week. In a 2005 São Paulo study of 455 babies, where the team measured two different angles later on strict side-view photos, 86% to 87% of the 169 at 12 weeks were called right. We found no published test of guesses by apps, paid services or online groups, and no scan is 100% sure.
Is nub theory accurate at 13 weeks?
In studies of the angle method, calls at 13 weeks were more accurate than at 11 weeks. In a 1999 London study of 172 pregnancies at 11 to 14 weeks, where researchers judged the angle during the scan, every call at 13 weeks was right; the abstract gives no count for that week. In a 2005 São Paulo study of 455 babies, where the team measured two different angles later on strict side-view photos, 87% to 88% of the 159 at 13 weeks were called right. No scan is 100% sure.
What does a boy nub look like compared with a girl nub?
Seen from the side, through the baby’s midline, a boy’s nub tends to point up toward the head, and a girl’s to lie flat or point down. Researchers draw a line through the skin of the lower back and measure the nub against it: over 30 degrees suggests a boy, and under 10, or parallel, a girl. Several studies make no call in between. These rules assume an exact side view, which a scan picture may not show.
Can nub theory be wrong?
Yes, especially early. In a 1999 London study of 172 pregnancies scanned at 11 to 14 weeks, more than half the boys at 11 weeks (56%) were called girls. Measurements also vary: in a São Paulo study, when two operators each photographed the same 30 babies, their angles differed by an average of 9 degrees for one angle and 15.7 for the other. No scan is 100% sure.
Are paid nub theory readings accurate?
We found no independent check. In October 2026, two of four paid services we looked at claimed 99.2% and 99.9% accuracy without saying how they counted, and all four said they weren’t medical professionals or didn’t give medical advice. Three said they don’t refund a guess that turns out wrong. A reading from a photo isn’t a medical result.
Can I use nub theory on my 12-week scan picture?
You can guess, but a print may not show what the method needs: an exact side view through the baby’s midline, with the lower back in view. If the baby is curled up or arched back, researchers warn, “the angle may be overestimated or underestimated and the result may be misleading.” For an early answer in the U.S., a blood test ordered by your doctor or midwife can report the sex from about 10 weeks, though it isn’t 100% accurate either.
What is the nub on an ultrasound?
It’s the genital tubercle, a small bump between the baby’s legs that boys and girls both have early on. It later becomes the penis or the clitoris. From about 12 weeks, a sonographer may see it pointing up toward the head, which suggests a boy, or lying flat or pointing down, which suggests a girl. No scan is 100% sure.
No. Tested on 2,840,755 Swedish births, the Chinese gender chart did no better than a coin toss. How it’s read, where it came from, and what works instead.
At-home kits look for a boy’s DNA in your blood. Makers claim 99% or more from 6 or 7 weeks (2026); a 2011 review found lab tests before 7 weeks unreliable.