Mammalian ovaries contain a large stock of oocytes enclosed in primordial follicles. Ovarian cyclic activity induces some of these follicles to initiate growth towards a possible ovulation. However, most of these follicles terminate their growth at any moment and degenerate through atresia. In growing follicles, only a subset of oocytes are capable to support meiosis, fertilization and early embryo development to the blastocyst stage, as shown through embryo in vitro production experiments. This proportion of competent oocytes is increasing along with follicular size. Growing lines of evidence suggest that oocyte competence relies on the storage of gene products (messenger RNA or protein) that will be determinant to support early stages of embryo development, before full activation of embryonic genome. Given these facts, the question is: are these gene products stored in oocytes during late folliculogenesis, allowing an increasing proportion of them to become competent? Alternatively, these transcripts may be stored during early folliculogenesis as the oocyte grows and displays high transcription activity. Several arguments support this latter hypothesis and are discussed in this review: (i) many attempts at prolonged culture of oocytes from antral follicles have failed to increase developmental competence, suggesting that developmental competence may be acquired before antral formation; (ii) the recent discovery of oocyte secreted factors and of their ability to regulate many parameters of surrounding somatic cells, possibly influencing the fate of follicles between ovulation or atresia, suggests a central role of oocyte quality in the success of folliculogenesis. Finally, in addition to their role in interfollicular regulation of ovulation rate, late folliculogenesis regulation and atresia could also be seen as a selective process aimed at the elimination through follicular atresia of oocytes that did not succeed to store proper gene products set during their growth.