Nub Theory: How Accurate Is It at 11, 12 and 13 Weeks?

What is nub theory?

Nub theory says you can tell a boy from a girl at the 12-week scan by the angle of the "nub" — the genital tubercle, a small bud of tissue between the legs that every baby has early on. In girls it develops into the clitoris; in boys it goes on to form part of the penis.

On a side-on view of the baby, the nub is compared with a horizontal line drawn through the skin of the lower back. In the original research:

• Boy: the nub points upwards at an angle of more than 30 degrees to that line.
• Girl: the nub lies parallel to the line or angles towards it.

"Nub theory" is the popular name; the method itself comes from research. Hospital specialists studied it for a practical reason: families at risk of a genetic condition that affects boys could decide earlier whether they needed an invasive test. That history matters, because every accuracy figure below comes from trained specialists, not from photos posted online.

How accurate is nub theory at 11, 12 and 13 weeks?

It depends heavily on the week. Every study that reported week by week shows the same pattern: unreliable at 11 weeks, much better at 12, best at 13.

* Whitlow counted scans where no call could be made as failures; by 14 weeks the figure was 90%. † Hsiao's figures are for babies where a call was possible; at 11 weeks, 40.6% of babies could not be assigned at all. ‡ Efrat 2006 left babies with an in-between angle unassigned, and made a call in 613 of 656.

Two studies give figures for the whole 11 to 14 week window. In Granada, Spain, a call was possible for 90.5% of 672 babies and correct for 87.5% of those. In a private obstetric practice in New South Wales, Australia, seven sonographers on routine scans got 75% right overall at 11 to 14 weeks, or 91% leaving out scans where they could not decide — but only 54% for babies under 12 weeks.

Is nub theory more accurate for boys or girls?

Early on, boys are the ones misread. In the 1999 London study, 56% of boys at 11 weeks were wrongly called girls, against 5% of girls wrongly called boys; by 12 weeks the boys' error rate was 3%. The angle in boys increased as they grew, so a boy's nub can look flat before it has had time to lift. The Australian study found the same: boys under 13 weeks were more likely to be called wrongly or not called at all. A flat-looking nub at 11 weeks is a weak sign of a girl.

Why it depends on the scan and the sonographer

Nub theory is not a yes-or-no sighting. It is a measurement, and measurements vary with the person taking them and the conditions.

• The sonographer. The Australian study found statistically significant differences between individual sonographers' success rates. A 2024 Saudi study of 312 babies found agreement on the nub angle between two operators was only moderate (an intraclass correlation of 0.695), lower than for a different measurement, the distance between the genitals and the anus (0.871).
• How far along you are. The Granada team found the best cut-offs for an accurate call were a baby measuring more than 55.7 mm from crown to rump and a pregnancy of more than 12 weeks and 2 days. Our crown-rump length calculator converts one to the other.
• Your body. In the same study, accuracy fell as the mother's body mass index rose.
• The view. The studies measured the angle on a true side-on view through the middle of the baby, which the sonographer adjusts in real time.

Can you use nub theory on a scan photo?

Not in the way the studies did. Every figure above comes from a trained operator looking at a live scan, choosing the view, judging an angle and, in most of the studies, being allowed to say "can't tell". We found no published study that tested people guessing from printed or phone photos of a scan, which is how nub theory is mostly used online. A photo is a single frame, and you cannot tell from it whether the view was truly side-on. Treat a nub guess from a picture as a fun guess, not a result.

What about nub theory at 9 or 10 weeks?

None of the clinical studies of routine scans that we found started before 11 weeks, and 11 weeks was the least accurate point in every one of them. After testing routine scans, the Australian authors advised that "predictions prior to 12 weeks should be discouraged".

A 2026 anatomy study from Bucharest, which dissected embryos and used 3D ultrasound, reported that the nub already points in different directions in boys and girls from as early as 6 weeks. Its abstract gives no accuracy figure for calling the sex on a scan, so it does not tell you how often an early guess would be right.

Other ways to find out the sex

Nub theory is the only early-scan method with published studies, but it is not the only route, and the 12-week scan has other jobs to do: see the 12-week scan. For when the sex is usually confirmed, and how reliable each route is, see when you can find out the sex. A blood test of the baby's DNA is the other early route: see NIPT and what it can tell you. Your baby's development at this stage is covered in 11 weeks pregnant, 12 weeks pregnant and 13 weeks pregnant.

If you are comparing early-scan theories, skull theory and Ramzi theory have no published studies at all, and the folklore methods are covered in old wives' tales for boy or girl.

Frequently asked questions

Is nub theory accurate at 12 weeks?

Fairly, in trained hands. Studies of specialists measuring the angle on a live scan found it right between 75% and 99% or more of the time at 12 weeks, depending on the study and on whether scans with no call counted as misses. It is more reliable at 13 weeks and much less so at 11.

Is nub theory ever wrong?

Yes, and most often early. At 11 weeks, one London study found more than half of boys were wrongly called girls, because a boy's nub can still look flat. Accuracy also varies between sonographers and falls with higher maternal body mass index. A guess from a scan photo has never been tested at all.

Can you do nub theory at 10 weeks?

Not with any evidence behind it. The clinical studies we found started at 11 weeks, and 11 weeks was the least accurate point in all of them, with correct calls between 46% and 72%. Australian researchers who tested routine scans advised that predictions before 12 weeks should be discouraged.