Does Nub Theory Work for Twins? A Realistic Guide

Yes. Nub theory uses exactly the same method for twins as for a single baby — there is nothing different about the underlying biology. Each twin develops a genital tubercle (the “nub”), and each one is read using the 30° rule on a clear mid-sagittal view.

The rule comes from the foundational first-trimester study by Efrat and colleagues (Efrat et al., 1999): a nub angled more than 30° from the lower body suggests a boy, while a flatter, parallel nub suggests a girl. The difference with twins is purely practical — getting a clean, separate view of each baby is harder. If you’re new to the method, start with our complete guide to nub theory.

The challenge with twins isn’t the theory — it’s the picture. Two babies sharing a small space simply give the sonographer fewer clean angles to work with.

• Overlap: one baby can lie in front of the other, hiding the nub.
• Crowding: less room to move means babies are more likely to be curled up or pressed together.
• Facing away: if a baby’s back or bottom is toward the probe, there may be no usable side-on profile.
• Telling them apart: each nub must be matched confidently to the correct baby.

Because of all this, confidence is often lower for twins than for a singleton — sometimes for one baby, sometimes for both. The honest accuracy with a clear image is around 85–90% at 12–13 weeks (Manzanares et al., 2016), and a twin scan that only partly cooperates can land below that. Anyone promising 99–100% for twins is overselling it.

The single most important thing is that each baby has its own clear mid-sagittal (side-on) view. One great image of one twin and a blurry image of the other means only one baby can be called confidently.

A strong twin scan for nub theory usually has:

• A separate side-on profile for each baby, not a single frame showing both at once.
• The whole lower body and nub visible, so the angle can actually be measured.
• Each baby clearly identifiable (e.g. labelled Twin A and Twin B by the clinic).
• Good image clarity, with the nub not hidden behind a limb, the cord, or the other baby.

If your scan photos show both babies but neither in a clean profile, it’s worth asking the sonographer for an individual side-on shot of each. For more on what to look for, see how to read a 12-week scan.

Each twin is treated as its own reading. We don’t average the babies or guess one from the other — Twin A and Twin B are assessed independently against the same threshold.

For each baby, on its own clear mid-sagittal view, the angle of the genital tubercle is measured relative to the lower body:

• Boy: nub angled more than 30° upward.
• Girl: nub parallel or under ~10°, lying flat along the lower body.
• Indeterminate: angles in between can’t be reliably assigned.

This means a twin scan can come back as two boys, two girls, or one of each — and each result carries its own confidence level depending on how clear that baby’s image was. The same 12–13 week timing and ~65 mm crown-rump length guidance applies to both babies (Efrat et al., 1999).

It is genuinely common for one twin to be readable and the other not — and that’s nothing to worry about. If a baby is facing away, tucked behind its sibling, or curled up, there may be no usable side-on view, and a confident reading simply isn’t possible from that scan.

When that happens, the honest answer is that one or both babies can’t be assigned yet. We won’t guess to fill the gap — a forced reading is worse than no reading. A later scan, once the babies have grown and shifted position, usually gives a clearer picture. Because nub theory is an early, for-fun estimate and not a diagnostic test, the 20-week anatomy scan remains the medical confirmation for both babies.

This is also where our money-back guarantee matters: if we can’t give you a confident reading from your scan, you’re covered.

The timing for twins is the same as for a single baby: aim for 12–13 weeks. Earlier than that, the genital tubercle hasn’t differentiated enough to read reliably for either baby.

• Before ~12 weeks: too early — wait until each baby is developed enough to read.
• 12–13 weeks: the ideal window, especially once each baby’s crown-rump length passes ~65 mm.
• If one baby is poorly positioned: it’s fine to submit anyway — we’ll call the baby we can and flag the one we can’t.
• If both babies are facing away or hidden: a slightly later or repeat scan often produces a much more readable image.

The summary: same method, same timing, just a higher bar for image quality. A clear side-on view of each baby is what turns a twin scan from “maybe” into a confident reading.

The method and timing are identical — what changes is how hard it is to get a clean image of each baby, and therefore how confident the reading is.

If you’d like the full picture on how reliable the method is, our guide to nub theory accuracy breaks down the numbers, and you can read about how our AI works alongside our human double-check.