Undifferentiated spermatozoa (spermatids) must undergo a striking change in nuclear structure and chromatin. The genetic outcomes of this chromatin remodeling in mammals are yet unknown. Our research team now investigates the nuclear events of these crucial differentiation steps in order to establish whether de novo variation (mutations) may be produced. We have established that the chromatin remodeling is associated with transient DNA double strand breaks (DSBs) in the whole population of differentiating spermatids. In such a haploid context, we hypothesize that DSB repair will induce mutations. When the subset of genes is being analyzed, our initial genome-wide mapping of DSB in mouse spermatids indicates that they arise preferentially within neurodevelopmental genes. Hence, any alteration in DNA repair during this transition could create functional mutations within these genes leading to neurological disorders. DSBs are highly conserved in lower eukaryotes allowing the yeast model to be used to decipher the molecular mechanism of strand breaks formation and repair. We are currently setting up a clinical test based on the specific quantification of DSBs as a direct determination of the genetic integrity of the male gamete and the father’s propensity to transmit neurological disorders to the next generation.