Can abnormal embryos self-correct?

This is an active area of scientific discussion rather than a settled question, and you should treat confident answers in either direction with caution. The observation behind it: embryos are mosaic more often than expected, and some pregnancies proceed normally after transfer of embryos with abnormal or mosaic results. Proposed explanations include abnormal cells being outcompeted by normal ones during development, or being preferentially allocated to the placenta rather than the baby. Whether this constitutes self-correction, or simply reflects that the biopsy did not represent the embryo accurately in the first place, is not resolved. ESHRE's position on the practical question is that the clinical significance of transferring mosaic embryos is currently unknown. What this does not mean: it does not mean embryos reported as fully abnormal should be transferred in the hope they will correct themselves. There is no basis for that. What it does mean: a mosaic or uncertain result is less definitive than it sounds, and the possibility that the tested cells did not represent the embryo is real. If you are being offered transfer of an embryo on the basis of possible self-correction, ask what specifically was found, what your clinic's documented policy is, and what evidence they are relying on. This is territory where enthusiasm sometimes runs ahead of data. Genetic counselling before deciding is worthwhile.