Boy Nub vs Girl Nub: How to Tell the Difference on a 12-Week Scan

The honest answer is that you read three things together: the angle of the nub, its shape, and whether it shows stacking. The angle does most of the work.

On a true mid-sagittal (side-profile) view, draw an imaginary horizontal line through the lower-back skin. If the nub lifts more than 30° above that line, it leans boy; if it lies parallel or under 10°, it leans girl (Efrat et al., 1999). Anything between 10° and 30° is genuinely indeterminate — that middle zone is real, and it is why a single glance is never a guarantee.

This page assumes you already know the basics. If you want the foundations first — what the nub actually is and where the 30° rule comes from — start with our complete guide to nub theory, then come back here to compare boy vs girl.

A boy nub has three giveaways. First, the angle: it tilts upward at more than 30° from the lower body, sometimes dramatically so by 13 weeks. Second, the shape: it tends to look thicker, brighter, and more defined than a girl nub. Third, the stacking sign: layered white lines stacking up behind the nub as the scrotum begins to form.

Think of a boy nub as standing away from the spine. The steeper and more defined it looks on a clear, properly side-on image, the more confident a boy call becomes — especially once the baby has grown past around 65 mm crown-rump length.

A girl nub is defined more by what it doesn’t do. The angle stays flat — parallel to the lower body, or under 10°. The shape is finer and flatter, without the bright, chunky definition of a boy nub. And it usually shows no stacking.

A girl nub often sits closer to the spine, hugging the line of the lower back rather than lifting away from it. The catch is that an early boy nub that hasn’t angled up yet can briefly look flat — which is exactly why female calls are the harder ones at 12 weeks and get more reliable into week 13.

The stacking sign is one of the strongest boy indicators. As a male fetus develops, the forming scrotum shows up on ultrasound as two or three short, layered white lines stacked behind the nub — a little like a tiny pile of stripes.

If you can clearly see stacking, it’s a meaningful nudge toward boy, because girl nubs rarely produce that layered look. But stacking is subtle, it needs a sharp image, and it’s easy to imagine on a blurry scan. Treat it as supporting evidence for the angle, not a standalone verdict — the nub is only a few pixels wide and overconfidence here is a classic trap.

Here is the quick reference. Read the three rows together rather than relying on any single one — agreement across angle, appearance, and stacking is what makes a confident call.

Note that none of these cues mean much before 12 weeks. For why the 12–13 week window matters so much and how to capture a clean side-on view, see how to read a 12-week scan.

Most wrong calls come from a handful of repeat errors. Watch for these:

• Confusing the umbilical cord for the nub. The cord can sit between the legs and mimic a raised “boy” nub — the most common false-boy mistake.
• Judging too early. Before 12 weeks both sexes look alike; at 11 weeks accuracy drops sharply, and before that there is nothing reliable to measure.
• Reading a blurry or low-resolution scan. The nub is only a few pixels; a soft image turns angle and stacking into guesswork.
• Misjudging the angle. If the baby is curled, tilted, or the view isn’t truly mid-sagittal, the measured angle is distorted and the call is unreliable.

Each of these is avoidable with the right image and a careful, consistent method — which is exactly where untrained home guessing tends to fall down.

The hardest part of boy-vs-girl isn’t knowing the rules — it’s applying them consistently to a tiny, noisy image. Two people can look at the same nub and disagree, and the angle is easy to misread when the baby is curled or the view isn’t perfectly side-on. Realistic accuracy is about 85–90% with a clear 12–13 week image even for experienced readers (Manzanares et al., 2016; 87.5% correct of assignable cases, n=672) — so anyone promising 99–100% certainty is overselling it.

That is the gap NubHub is built to close. Our AI checks angle, shape, stacking, and image clarity together instead of fixating on one cue, then a real person double-checks the result before it reaches you — typically within 2 hours. For privacy and consent reasons we don’t publish real labelled scans, so the most reliable way to read your nub is to let our AI analyse it rather than squint at it alone.

Want to know more before you upload? Compare confidence levels in our nub theory accuracy guide, or browse common questions on the FAQ.