> And presumably if both parents have already done it why would you need to test the baby?

Before all the privacy concerns, I was considering getting my kids tested, even though my wife and I have both done it.

Mainly so that I could see which genes they picked up from each of us and if they got any new mutations.

Also, a situation that might be somewhat niche -- we had IVF, so it is possible there was a mistake in the clinic and the child wasn't made from us. I read too many stories about IVF clinics where the samples were switched by the employees with their own.

Even more mind blowing, through sheer clerical error, the children that came out of my wife could be mine but not hers!

Given that they look like us we're pretty sure the clinic didn't screw up, and at this point I would keep them regardless, but if I could discover an IVF mistake, the settlement would make for one heck of a college fund.

It would be extraordinarily rude . But it's definitely not what the parent comment meant. :)

Mutation is a thing.

"No, no, I wasn't questioning the paternity, you can use the test to check if your baby is a mutant."

Given the popularity of the X-men i think that might improve sales enough to reverse the trend.

But 23andme will always show some differences, while individual SNPs have high accuracy >99% due to the amount that is being tested you will always have some results that are wrong. Multiple testing problem is an issue here.