The nub theory is a method of guessing a baby’s sex during early pregnancy, typically at the 11- to 13-week ultrasound. It works by looking at the angle of a small structure called the genital tubercle, a proto-genital organ that all fetuses develop before the external sex organs fully form. The idea is simple: if the nub points upward, it’s more likely a boy; if it stays flat or angles downward, it’s more likely a girl. It’s not a medical diagnosis, but it has some real anatomy behind it.
What the “Nub” Actually Is
Early in pregnancy, every fetus develops a genital tubercle, a tiny bud of tissue between the legs. This structure is the precursor to both male and female genitalia. The clitoris and labia become identifiable as early as 7 weeks of development, but at that stage everything still looks remarkably similar on ultrasound regardless of sex. By about 9 weeks, the tip of the clitoris separates into its own structure, and by 10 weeks a skin fold begins forming around it. In male fetuses, the same tissue develops into the penis.
Because male and female anatomy starts from the same structure and diverges gradually, the genital tubercle looks nearly identical on ultrasound at 11 weeks. This is why the theory relies not on shape but on angle. The further along the pregnancy, the more the tubercle has committed to its final form, and the easier it becomes to read.
How the Angle Prediction Works
The nub theory requires a side-on (sagittal) ultrasound image where you can see the baby’s full spine and the genital area in profile. From that view, you measure the angle of the genital tubercle relative to the lower spine.
- 30 degrees or more above the spine: The nub is pointing upward, which supposedly indicates a boy.
- 10 degrees or less: The nub is roughly parallel to the spine or angled slightly downward, which supposedly indicates a girl.
- Between 10 and 30 degrees: The angle falls in a gray zone and doesn’t give a clear prediction either way.
Another way radiologists describe this: if the midline sagittal view shows a “cranial notch” (the nub tilting toward the head), the fetus is male. If it shows a “caudal notch” (tilting toward the tail end), the fetus is female. These are just different ways of describing the same directional difference.
How Accurate It Really Is
Accuracy depends heavily on timing. A study published in the Australasian Journal of Ultrasound in Medicine tracked predictions made by sonographers during routine scans and found a clear pattern: the later the scan, the better the guess.
At 11 weeks, predictions were correct only 54% of the time, which is essentially a coin flip. By 12 weeks, accuracy jumped to 77%. At 13 weeks, it reached 79%. And from 14 weeks onward, every prediction in the study was correct, though the sample size at that stage was small. A separate analysis found the method reaches 100% accuracy after 13 weeks when performed by an experienced practitioner using proper imaging technique.
The takeaway: if you’re trying to apply nub theory at your 12-week scan, you have roughly a three-in-four chance of getting the right answer. At 11 weeks, you might as well guess.
Why Predictions Go Wrong
Several practical factors make early nub readings unreliable. The baby’s position in the womb matters enormously. If the legs are crossed, the umbilical cord is draped between the legs, or the baby is facing the wrong direction, you simply can’t get the profile view needed to measure the angle. Fetal movement is another problem: babies at 11 to 13 weeks are often highly active, making it hard to capture a clean image.
Even with a good image, the fundamental issue is biological. At 11 weeks, the genital tubercle hasn’t differentiated enough for the angle to be consistently different between sexes. The structure is still developing, and many female fetuses have a slightly elevated nub that can be mistaken for a male angle. This is why accuracy improves week by week as the anatomy becomes more distinct.
Image quality also plays a role. The resolution of the ultrasound machine, the skill of the sonographer, and even maternal body composition can affect how clearly the nub appears. Most sonographers won’t attempt a sex prediction at a first-trimester scan for exactly these reasons.
How It Compares to Other Methods
The standard clinical method for determining sex on ultrasound is a second-trimester anatomy scan, typically done around 18 to 20 weeks. By that point, the penis and scrotum or the labia are clearly visible, and accuracy is very high without needing to measure angles.
For earlier results, cell-free DNA blood tests (often called NIPT) can identify fetal sex as early as 10 weeks with accuracy above 99%. These tests analyze fragments of fetal DNA circulating in the mother’s blood and are primarily used to screen for chromosomal conditions, with sex determination as a secondary benefit. If knowing the sex early is important to you for medical reasons, such as screening for sex-linked genetic disorders, a blood test is far more reliable than any ultrasound-based method in the first trimester.
The nub theory occupies a middle ground: it’s free, it uses an ultrasound you’re already getting, and it can be fun to try. But it’s not a clinical tool, and no major radiology guideline recommends using it for definitive sex determination before 14 weeks.
Trying It on Your Own Scan
If you want to take a look at your own 12- or 13-week ultrasound image, you need a clear side profile of the baby where the spine runs roughly horizontally across the image. Look between the legs for a small protrusion. If it angles noticeably upward from the body, nub theory would call that a boy. If it runs along the same line as the lower spine or dips slightly, it would call that a girl.
In practice, most parents find it genuinely difficult to identify the nub on their own scan photos. The images are small, the anatomy is tiny, and the angle difference between 20 degrees and 35 degrees is not obvious to an untrained eye. Online communities dedicated to nub theory often have experienced members who review scan photos, but even those readings are informal guesses. If the angle falls in the ambiguous 10-to-30-degree range, no amount of squinting at the image will give you a reliable answer.