DNA methylation is a major epigenetic modification of vertebrate genomes, mainly found in the CpG dinucleotide context. DNA methylation in the non-CpG context (mCH, H=A, C or T) has also been identified in the vertebrate subphylum, however, at much lower quantities and in a highly tissue-restricted manner. One of the major impediments of studying the function of mCH in mammals is the fact that it cannot be uncoupled from mCG, as they are both written by DNMT3A/B. Our lab has recently described a peculiar mCH reprogramming event that takes place at mosaic satellite repeats (MOSAT) in teleost genomes. Additionally, it was discovered that in zebrafish Dnmt3ba was specifically responsible for this form of mCH and its loss of function did not affect mCG. This offers a unique opportunity to study the function of mCH independent of mCG. Here we investigate to what extent mCH reprogramming at MOSAT repeats plays a role in early zebrafish embryonic development. By using base-resolution DNA methylome quantification, TALE-based imaging, as well as transcriptome profiling, in diverse zebrafish loss- and gain of function models, we are working towards understanding the function of this mark. To date we have observed an increase in early embryo mortality and altered transcriptomes in zebrafish models that are deficient in mCH at MOSAT repeats. In addition, imaging using MOSAT specific TALE’s shows a change in DNA localisation in early mutant embryos when compared to WT, where mCH deficient mutants exhibit reduced localisation of MOSAT repeats to the nuclear periphery in early developmental stages. Finally, we have identified an inverse relationship between mCH and H3K9me3 at a subset of MOSAT repeats, suggesting mCH may have a structural role in early embryonic development. Together these results contribute to establishing completely novel functions for one of the oldest, and best described gene-regulatory marks.