Gender Prediction Apps: Do Any Work?
The App Store and Google Play are full of gender prediction apps promising to tell you whether you're having a boy or a girl. Some use AI, some use ancient charts, and some ask about your cravings. With so many options, it's natural to wonder whether any of them actually have predictive power — or whether you'd be just as well off flipping a coin.
Table of Contents
Nub Theory Apps
Nub theory is the practice of predicting a baby's sex by analysing the angle of the genital tubercle (the "nub") visible on a first trimester ultrasound scan. In the hands of an experienced, trained reader looking at a good quality scan image, it can be a reasonably reliable method — particularly from 13 weeks.Several apps now claim to replicate this process using artificial intelligence, asking you to upload your scan image and returning a boy or girl prediction.
The honest assessment: app-based nub theory analysis varies enormously in quality. The best apps apply image analysis to the scan in a methodical way; the worst are essentially returning a random result dressed up in confident language. No app currently available to consumers matches the accuracy of an experienced human reader who knows how to identify a clear nub, assess the angle correctly, and recognise when a scan image is not suitable for a reliable read.
A well-trained human reader will also tell you when a prediction cannot be made — either because the nub is not clearly visible, the angle is ambiguous, or the scan is too early. Many apps do not do this, and will offer a confident prediction regardless of image quality.
If you're using a nub theory app, treat the result as a fun piece of entertainment rather than reliable information.
Old Wives' Tales Apps
A large category of gender prediction apps are built around traditional folklore methods: the Chinese gender calendar (which uses your age and the month of conception), heart rate prediction (the idea that girls have faster fetal heart rates), craving-based guessing, and the ring-on-a-string test, among others.These methods have been tested informally many times over many generations, and the conclusion is consistent: they are no more accurate than chance. An app that combines several of these methods is combining several sources of noise — the result is no more meaningful than a single coin toss.
They can be genuinely fun to use as a way of building excitement about your baby, and there is nothing wrong with enjoying them in that spirit. Just don't plan your nursery decor around the result.
What Actually Works
If you want a reliable answer about your baby's sex, these are the methods with genuine predictive power:NIPT (Non-Invasive Prenatal Testing). A blood test taken from 10 weeks of pregnancy that analyses fetal DNA in the mother's bloodstream. When it comes to sex determination, NIPT is highly accurate. It is available privately in the UK from around 10 weeks and is offered on the NHS in certain circumstances. It also screens for chromosomal conditions.
Nub theory by an experienced reader. At 13 weeks, a clear scan image read by someone trained and experienced in nub theory can be a reliable predictor. Results are less reliable before 12 weeks or from unclear images. This is offered as a private service by a number of specialists.
The 20-week anatomy scan (anomaly scan). Your NHS anomaly scan, offered around 18–21 weeks, can usually determine sex visually when the baby is in a suitable position. This scan is primarily about checking the baby's development, but most sonographers will tell you the sex if you ask and the position allows it.
SneakPeek and similar at-home DNA tests. These at-home blood spot tests claim to detect fetal DNA and determine sex from as early as 8 weeks. When used exactly as directed, their stated accuracy is high for detecting the Y chromosome, but results can be affected by contamination. They are not NHS-approved but are widely used privately.
Frequently Asked Questions
### Do AI nub theory apps work?Some AI nub theory apps are more systematic than others, but none currently match the reliability of an experienced human reader. The accuracy of the result depends heavily on the quality of your scan image, the clarity of the nub, and the sophistication of the app's analysis. Treat any app-based nub theory prediction as entertainment rather than fact, particularly if the app doesn't tell you when an image is unsuitable for a reliable read.
What's the most reliable way to predict gender before the 20-week scan?
NIPT blood testing (available privately from 10 weeks) is highly accurate for sex determination. Nub theory at 13 weeks by a trained reader is a good option for those who want an earlier, less medicalised prediction. Apps and old wives' tale methods are approximately no better than chance and should be treated as fun, not information.
Can old wives' tales really predict gender?
No — not reliably. Methods like the Chinese gender calendar, heart rate theory, cravings, and the ring test have no evidence base for gender prediction. In any individual case they will sometimes be right (approximately half the time), but this is the expected outcome when guessing between two options, not evidence that the method works. Use them for fun; don't rely on them for decisions.
For any questions about your pregnancy or antenatal care, your midwife or GP is always your best source of guidance.
This article is for informational purposes only and does not constitute medical advice.
Related articles: